CN108229499A - Certificate recognition methods and device, electronic equipment and storage medium - Google Patents
Certificate recognition methods and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN108229499A CN108229499A CN201711052744.8A CN201711052744A CN108229499A CN 108229499 A CN108229499 A CN 108229499A CN 201711052744 A CN201711052744 A CN 201711052744A CN 108229499 A CN108229499 A CN 108229499A
- Authority
- CN
- China
- Prior art keywords
- certificate
- image
- detected
- information
- certificate image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 73
- 238000013528 artificial neural network Methods 0.000 claims abstract description 71
- 238000000605 extraction Methods 0.000 claims abstract description 28
- 238000012549 training Methods 0.000 claims description 55
- 238000012545 processing Methods 0.000 claims description 33
- 238000004590 computer program Methods 0.000 claims description 10
- 238000005516 engineering process Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 description 13
- 238000004891 communication Methods 0.000 description 11
- 238000001514 detection method Methods 0.000 description 8
- 238000009499 grossing Methods 0.000 description 8
- 230000002708 enhancing effect Effects 0.000 description 6
- 239000002131 composite material Substances 0.000 description 5
- 239000000284 extract Substances 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000004069 differentiation Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000010191 image analysis Methods 0.000 description 2
- 238000003709 image segmentation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005242 forging Methods 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/513—Sparse representations
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The embodiment of the invention discloses a kind of certificate recognition methods and device, electronic equipment and storage medium, wherein method includes：Obtain certificate image to be detected；Certificate image to be detected is handled, obtains image to be detected region for including the certificate to be detected；Feature extraction is carried out to image to be detected region according to deep neural network, obtains comprehensive characteristics；Certificate image is carried out according to the comprehensive characteristics and forges identification, obtains the recognition result of the certificate image.The embodiment of the present invention comprehensive and accurate can identify forged certificate, without carrying out the addition of additional information or equipment, implement easier.
Description
Technical Field
The present invention relates to image processing technologies, and in particular, to a certificate identification method and apparatus, an electronic device, and a storage medium.
Background
Certificate anti-counterfeiting is an important problem in the field of anti-counterfeiting technology, and has important application in many fields, especially in the field of internet finance, and the certificate anti-counterfeiting technology can effectively protect account security of an account holder.
Disclosure of Invention
The embodiment of the invention provides a technical scheme for certificate identification.
The certificate identification method provided by the embodiment of the invention comprises the following steps: acquiring a certificate image to be detected; processing a certificate image to be detected to obtain an image area to be detected containing the certificate to be detected; extracting the characteristics of the image area to be detected according to the deep neural network to obtain comprehensive characteristics; and carrying out certificate image forgery identification according to the comprehensive characteristics to obtain an identification result of the certificate image.
In an alternative, the method further comprises: training the deep neural network; the training the deep neural network comprises: collecting each certificate image sample, wherein the certificate image sample is marked with information whether to be forged or not; extracting the features of the certificate image sample according to the deep neural network to obtain the comprehensive features of the certificate image sample; carrying out certificate forgery identification according to the comprehensive characteristics to obtain an identification result of the certificate image sample; and training the deep neural network based on the identification result of the certificate image sample and the marked fake information.
In an optional mode, when the certificate image sample is a certificate image forged sample, the certificate image forged sample is also marked with forged source marking information; the training the deep neural network further comprises: acquiring counterfeiting source prediction information of a certificate image counterfeiting sample according to the comprehensive characteristics; and training the deep neural network based on the forged source prediction information and the forged source marking information.
In an alternative, the training the deep neural network further includes: and when a certificate image forgery new sample is acquired, training the deep neural network based on the certificate image forgery new sample, wherein the certificate image forgery new sample is obtained by a certificate image forgery technology.
In one alternative, the composite features include: at least one or more of local binary pattern features, sparsely encoded histogram features, color features, panorama features, document region features, and document detail features.
In an optional mode, the local binary pattern feature is used for representing edge information in an image to be detected; the sparse coded histogram features are used for representing reflection and fuzzy information in an image to be detected; the color features are used for distinguishing the certificate which is really shot from the certificate copy; the panoramic image characteristics are used for extracting counterfeiting information of the image to be detected; the certificate area is used for determining a certificate area cutting chart in the image to be detected; and the certificate detail characteristics are used for extracting the counterfeiting information of the image to be detected.
In an alternative, the identification result of the certificate image includes: whether the certificate image is counterfeit or not; when the identification result is that the certificate image is forged, the method further comprises the following steps: and outputting a counterfeit source of the certificate image according to the comprehensive characteristics.
In an alternative, the method further comprises: and training the deep neural network according to the counterfeit identification result and the counterfeit source of the certificate image.
In an alternative, the document image is the document image itself or a video frame in the document video.
In an optional manner, the processing the document image to be detected to obtain an image to be detected includes: denoising, enhancing, smoothing and sharpening the certificate image to obtain the image area to be detected; or, performing video frame selection operation on the certificate video to obtain a target video frame image, and performing denoising, enhancing, smoothing and sharpening on the target video frame image to obtain the to-be-detected image area.
In an optional mode, the performing a video frame selection operation on the certificate video to obtain a target video frame image includes: aiming at the video frame of the certificate video, any one or any combination of the following operations is carried out: judging whether the certificate is positioned in the center of the video frame image, whether the edge of the certificate is completely included in the video frame image, whether the area proportion of the surface of the certificate in the video frame image meets a preset proportion, whether the definition of the video frame image meets a preset definition threshold value, and whether the exposure of the image in the video meets a preset exposure threshold value; and selecting the target video frame image from the certificate video according to the judgment result.
In an optional mode, for the certificate video to be detected, the method further comprises the following steps: if the identification results of all video frame images in the certificate video are not forged, determining that the certificate video is not forged; or if the identification result of any video frame image is that forgery exists, determining that the forgery exists in the certificate video.
In an optional mode, after obtaining the identification result of the certificate image, the method further includes: if no certificate image forgery exists, extracting certificate information, and comparing the extracted certificate information with information in a certificate library; responding to the extracted certificate information to be matched with the information in the certificate library, and requesting by a user; or responding to the fact that the extracted certificate information is not matched with the information in the certificate library, not requesting by a user, marking the certificate image as a forged certificate image, and training the deep neural network.
The embodiment of the invention provides a certificate recognition device, which comprises: the acquisition unit is used for acquiring a certificate image to be detected; the processing unit is used for processing a certificate image to be detected to obtain an image area to be detected containing the certificate to be detected; the extraction unit is used for extracting the characteristics of the image area to be detected according to the deep neural network to obtain comprehensive characteristics; and the identification unit is used for carrying out certificate image forgery identification according to the comprehensive characteristics to obtain the identification result of the certificate image.
In an alternative, the method further comprises: a training unit for training the deep neural network, the training unit comprising: the acquisition subunit is used for acquiring each certificate image sample, and the certificate image sample is marked with information whether to be forged or not; the extraction subunit is used for extracting the features of the certificate image sample according to the deep neural network to obtain the comprehensive features of the certificate image sample; the identification subunit is used for carrying out certificate forgery identification according to the comprehensive characteristics to obtain an identification result of the certificate image sample; and the training execution subunit is used for training the deep neural network based on the identification result of the certificate image sample and the marked fake information.
In an optional mode, when the certificate image sample is a certificate image forged sample, the certificate image forged sample is also marked with forged source marking information; the training unit further comprises: the counterfeit source acquisition subunit is used for acquiring counterfeit source prediction information of the certificate image counterfeit sample according to the comprehensive characteristics; and the fake training subunit is used for training the deep neural network based on the fake source prediction information and the fake source labeling information.
In an alternative, the training unit further comprises: and the newly-added sample training subunit is used for training the deep neural network based on the certificate image forged newly-added sample when the certificate image forged newly-added sample is acquired, and the certificate image forged newly-added sample is obtained through a certificate image newly-added forging technology.
In one alternative, the composite features include: at least one or more of local binary pattern features, sparsely encoded histogram features, color features, panorama features, document region features, and document detail features.
In an optional mode, the local binary pattern feature is used for representing edge information in an image to be detected; the sparse coded histogram features are used for representing reflection and fuzzy information in an image to be detected; the color features are used for distinguishing the certificate which is really shot from the certificate copy; the panoramic image characteristics are used for extracting counterfeiting information of the image to be detected; the certificate area is used for determining a certificate area cutting chart in the image to be detected; and the certificate detail characteristics are used for extracting the counterfeiting information of the image to be detected.
In an alternative, the identification result of the certificate image includes: whether the certificate image is counterfeit or not; when the identification result is that the certificate image is forged, the device further comprises: and the counterfeit source output unit is used for outputting the counterfeit source of the certificate image according to the comprehensive characteristics.
In an alternative, the apparatus further comprises: and the result training unit is used for training the deep neural network according to the counterfeit identification result of the certificate image and the counterfeit source.
In an alternative, the document image is the document image itself or a video frame in the document video.
In an alternative, the processing unit comprises: the certificate image processing subunit is used for carrying out denoising, enhancing, smoothing and sharpening on the certificate image to obtain the to-be-detected image area; or the certificate video processing subunit is configured to perform video frame selection on the certificate video to obtain a target video frame image, and perform denoising, enhancement, smoothing and sharpening on the target video frame image to obtain the to-be-detected image region.
In an optional manner, the credential video processing subunit is specifically configured to: aiming at the video frame of the certificate video, any one or any combination of the following operations is carried out: judging whether the certificate is positioned in the center of the video frame image, whether the edge of the certificate is completely included in the video frame image, whether the area proportion of the surface of the certificate in the video frame image meets a preset proportion, whether the definition of the video frame image meets a preset definition threshold value, and whether the exposure of the image in the video meets a preset exposure threshold value; and selecting the target video frame image from the certificate video according to the judgment result.
In an optional mode, for the certificate video to be detected, the method further comprises the following steps: a forgery determination subunit, configured to determine that there is no forgery in the certificate video if there is no forgery in the identification results for all the video frame images in the certificate video; or if the identification result of any video frame image is that forgery exists, determining that the forgery exists in the certificate video.
In an optional mode, after obtaining the identification result of the certificate image, the method further includes: the comparison unit is used for extracting certificate information if certificate image forgery does not exist, and comparing the extracted certificate information with information in a certificate library; the response unit is used for responding to the matching of the extracted certificate information and the information in the certificate library and requesting by a user; or responding to the fact that the extracted certificate information is not matched with the information in the certificate library, not requesting by a user, marking the certificate image as a forged certificate image, and training the deep neural network.
An embodiment of the invention provides a computer-readable storage medium, on which a computer program is stored, which, when executed by a processor, implements the steps of the certificate identification method.
The electronic equipment provided by the embodiment of the invention comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor realizes the steps of the certificate identification method when executing the program.
Based on the certificate identification method and device, the electronic equipment and the storage medium provided by the embodiment of the invention, compared with the prior art that certificate forgery identification needs to be carried out by means of extra information or special hardware equipment, the certificate image to be detected is subjected to feature extraction to obtain the feature information of a plurality of categories, so that the certificate forgery identification is carried out according to the comprehensive feature information of the plurality of categories, the forged certificate can be comprehensively and accurately identified, the addition of extra information or equipment is not needed, and the implementation is simpler and more convenient. In an optional mode, modeling is performed through strong description capacity of the deep neural network, so that the discrimination of the forged characteristic information is more accurate, and the real identity information and the forged identity information can be better distinguished; and through the differentiation of the counterfeit certificate source, the counterfeit information source can be used as an additional counterfeit type label to be provided for the service port for use.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description, serve to explain the principles of the invention.
The invention will be more clearly understood from the following detailed description, taken with reference to the accompanying drawings, in which:
FIG. 1 is a flow chart of one embodiment of a method of credential identification of the present invention.
FIG. 2 is a flow chart of another embodiment of a method of document identification of the present invention.
FIG. 3 is a flowchart of training a deep neural network according to an embodiment of the present invention.
FIG. 4 is a schematic view of an embodiment of a document identification device of the present invention.
FIG. 5 is a schematic view of another embodiment of a document identification device of the present invention.
Fig. 6 is a schematic structural diagram of an embodiment of an electronic device according to the present invention.
Detailed Description
Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that: the relative arrangement of the components and steps, the numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present invention unless specifically stated otherwise.
Meanwhile, it should be understood that the sizes of the respective portions shown in the drawings are not drawn in an actual proportional relationship for the convenience of description.
The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.
Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail but are intended to be part of the specification where appropriate.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, further discussion thereof is not required in subsequent figures.
The certificate anti-counterfeiting technology has important application in many fields in life, and particularly in the field of internet finance, the certificate anti-counterfeiting technology can effectively protect the safety of account opening people. Modeling of credential forgery information is however complex. On the one hand, the forged information has limited access, such as in the internet field, and the forged information is only from pictures or video files and cannot be distinguished according to the additional information of the holder. On the other hand, the characteristics of the certificates are different according to different sources of the certificates, and the complexity of the characteristics means that the modeling is more difficult. In the process of implementing the invention, the inventor finds that the existing certificate anti-counterfeiting method needs to use a lot of additional information or special hardware equipment, and along with the development of internet finance, a more convenient and efficient certificate anti-counterfeiting identification scheme needs to be mastered urgently.
FIG. 1 is a flow chart of one embodiment of a method of credential identification of the present invention. As shown in FIG. 1, the method of this embodiment includes S101-S104.
S101: and acquiring a certificate image to be detected.
In services such as internet finance, user real-name authentication is required. And the user shoots the uploaded certificate picture or video through the mobile terminal so as to further determine the identity of the user. For example, in a stock account opening operation performed by a user, the user is required to upload an identification card photo. At this time, the user uploads the picture after shooting through a terminal such as a mobile phone.
S102: and processing the certificate image to be detected to obtain an image area to be detected containing the certificate to be detected.
After the certificate image to be detected is obtained, processing is needed, so that the image area to be detected of the certificate to be detected, which meets the certificate forgery identification requirement, is obtained.
In the embodiment of the invention, the certificate image can be the certificate image (certificate picture) or a video frame in the certificate video.
And for the certificate picture to be detected, performing image processing such as denoising, enhancing, smoothing and sharpening on the certificate picture to obtain an image area to be detected. The image processing is the processing performed before feature extraction, segmentation and matching of an input image in image analysis, and the main purpose of the image processing is to eliminate irrelevant information in the image, recover useful real information, enhance the detectability of relevant information and simplify data to the maximum extent, thereby improving the reliability of feature extraction, image segmentation, matching and recognition.
For a certificate video of a certificate to be detected, firstly, carrying out video frame selection operation on the certificate video to obtain a target video frame image; and then, carrying out image processing such as denoising, enhancing, smoothing and sharpening on the target video frame image to obtain an image area to be detected. Similarly, image processing, which is processing performed before feature extraction, segmentation and matching of an input image in image analysis, is mainly intended to remove irrelevant information from an image, recover useful real information, enhance the detectability of relevant information, and simplify data to the maximum extent, thereby improving the reliability of feature extraction, image segmentation, matching and recognition.
The criteria for the video frame selection operation may include the following items: whether the certificate is positioned in the center of the image, whether the edge of the certificate is completely included in the image, the proportion of the image area occupied by the surface of the certificate, the image definition and the image exposure. By selecting frames from the video according to the above criteria, an image with satisfactory quality can be selected, for example, the selected image may be: the certificate is located in the center of the image, the edge of the certificate is completely included in the image, the certificate accounts for about 1/2-3/4 of the image, and the image is high in definition and exposure.
Therefore, the specific process of performing video frame selection operation on the certificate video to obtain the target video frame image may include the following steps: (1) aiming at the video frame of the certificate video, any one or any combination of the following operations is carried out: judging whether the certificate is positioned in the center of the video frame image, whether the edge of the certificate is completely included in the video frame image, whether the area proportion of the surface of the certificate in the video frame image meets a preset proportion, whether the definition of the video frame image meets a preset definition threshold value, and whether the exposure of the image in the video meets a preset exposure threshold value; (2) and selecting the target video frame image from the certificate video according to the judgment result.
And for the certificate video of the certificate to be detected: if the identification results aiming at all the video frame images in the certificate video are not forged, determining that the certificate video is not forged; or if the identification result of any video frame image is that the certificate video is forged, determining that the certificate video is forged.
S103: and extracting the features of the image area to be detected according to the deep neural network to obtain comprehensive features.
In the embodiment of the present invention, the comprehensive feature may be understood as a feature including a plurality of categories, where each category characteristic is used to represent a certain characteristic of the image region to be detected. For example, composite features include, but are not limited to: local binary pattern characteristics, sparsely encoded histogram characteristics, color characteristics, panorama characteristics, certificate region characteristics, and certificate detail characteristics. The more the types of the characteristic information are, the more comprehensive the counterfeit identification can be performed. Wherein: the local binary pattern features are used for representing edge information in the image to be detected; the sparsely encoded histogram features are used for representing reflection and fuzzy information in an image to be detected; the color features are used for distinguishing the certificate which is really shot from the certificate copy; the panoramic image characteristics are used for extracting the counterfeiting information of the image to be detected; the certificate area is used for determining a certificate area cutting chart in the image to be detected; the certificate detail features are used for extracting the counterfeiting information of the image to be detected.
S104: and carrying out certificate image forgery identification according to the comprehensive characteristics to obtain an identification result of the certificate image.
In the embodiment of the present invention, the forgery includes two cases: certificate counterfeiting; and (5) image counterfeiting. Certificate forgery refers to identifying the characteristics of the certificate in an image and judging whether the certificate is forged or not; image forgery is the process of recognizing an image, judging whether the image is an image obtained by subjecting another image to some processing (e.g., copying, PS, copying), and judging whether the image is forged.
In an alternative, the identification result of the certificate image includes: whether the certificate image is counterfeit or not; when the identification result is certificate image forgery, the method further comprises the following steps: and outputting the counterfeit source of the certificate image according to the comprehensive characteristics. Wherein sources of counterfeiting include, but are not limited to: copies, PS (photoshop editing), photographs, etc. When the identification result is the forgery of the certificate image, the method can further comprise the following steps: and training the deep neural network according to the counterfeit identification result and the counterfeit source of the certificate image (the specific way of training the deep neural network can be seen in steps S2011-S2014 below).
Compared with the prior art that the certificate counterfeiting identification needs to be carried out by means of extra information or special hardware equipment, the certificate counterfeiting identification method and the certificate counterfeiting identification device have the advantages that the characteristic extraction is carried out on the certificate image to be detected to obtain the characteristic information of multiple categories, so that the certificate counterfeiting identification is carried out according to the characteristic information of the multiple categories, the counterfeit certificate can be identified comprehensively and accurately, the extra information or the equipment does not need to be added, and the implementation is simpler and more convenient. Moreover, the fake certificates of various fake sources can be effectively identified through the strong description capability of the deep neural network.
FIG. 2 is a flow chart of another embodiment of a method of document identification of the present invention. On the basis of the embodiment of fig. 1, the embodiment of fig. 2 introduces the operation of performing feature extraction and certificate forgery identification on the certificate image to be detected by using a deep neural network as a certificate forgery feature extraction and identification model. The implementation process of this embodiment can be understood simply as follows: the method comprises the steps of determining a counterfeit feature information set (training model) in advance, extracting features of an image to be detected, judging whether the image to be detected contains feature information in the counterfeit feature information set, and determining whether the certificate is a counterfeit certificate.
Deep learning forms a more abstract class or feature of high-level representation properties by combining low-level features to discover a distributed feature representation of the data. The neural network is provided with an input layer, an output layer and a hidden layer, the input feature vectors reach the output layer through the transformation of the hidden layer, and classification results are obtained on the output layer. Deep Neural Networks (DNNs) can be understood as Neural Networks with Deep learning capabilities. For example, the modeling process of the deep neural network is that firstly, data is prepared, sample data to be classified is stored in a matrix in the form of each column, and the class labels of each sample are stored as a row of vectors in the data sequence; secondly, network configuration, parameter initialization and conversion are carried out; then, writing a loss function; and finally, training the sample data.
As shown in FIG. 2, the method of this embodiment includes S201-S204.
S201: and obtaining a certificate forged feature extraction and identification model through training the deep neural network.
In an alternative manner, referring to fig. 3, which is a flowchart of training a deep neural network in an embodiment of the present invention, the deep neural network is trained through the following steps S2011-S2014.
S2011: and collecting each certificate image sample, wherein the certificate image sample is marked with information whether to be forged or not.
For example, a manually labeled document image dataset (database) can be pre-established in which document images for various types of documents are stored. During the modeling process, an image sample is taken from the dataset.
The counterfeit information is a feature that reflects that the image is likely to be counterfeit through various features of the image, and is also called a counterfeit cue. For example, counterfeit information includes, but is not limited to: the method comprises the steps of reflecting light and fuzzy information in an image through HSC characteristics, obviously distinguishing certificates and certificate copy information which are really shot through RGB characteristics, extracting most obvious fake information (hack) in the image through LARGE, extracting fake information such as edges of moire fringes of a reflection and reproduction device screen based on SMALL characteristics, and extracting fake information such as PS (photoshop edition), moire of the reproduction screen, reflection and the like through TINY characteristics.
S2012: and carrying out feature extraction on the certificate image sample according to the deep neural network to obtain the comprehensive features of the certificate image sample.
The composite features include, but are not limited to, any of the following: local Binary Pattern (LBP) features, sparsely coded Histogram of Sparse Code (HSC) features, color (RedGreenBlue, RGB) features, panorama (LARGE) features, document area (SMALL) features, document detail (TINY) features.
The edge information in the image to be detected can be highlighted through the LBP characteristics; by means of HSC characteristics, reflection and fuzzy information in an image to be detected can be reflected more obviously; by means of RGB characteristics, the really shot certificate and the certificate copy information can be distinguished more obviously; the LARGE features are full-image features, and most obvious fake information (hack) in the image to be detected can be extracted based on the LARGE features; the certificate area (SMALL) is an area tangent image with a plurality of times of the certificate area (for example, 1.5 times of the certificate area) in the image to be detected, comprises the part of the certificate, the certificate and the background, and can extract forged information such as the edge of Moire patterns on a screen of a reflecting and copying device based on the SMALL characteristics; the certificate detail drawing (TINY) is a cut-out drawing of a region with the size of a certificate-taking component, contains the certificate, and can extract counterfeit information such as PS (photoshop edition), Moire patterns of a reproduction screen, light reflection and the like based on the TINY characteristics.
S2013: and carrying out certificate forgery identification according to the comprehensive characteristics to obtain an identification result of the certificate image sample.
S2014: and training the deep neural network based on the identification result of the certificate image sample and the marked fake information.
The fake information of the fake certificate contained in the characteristics can be learned by the deep neural network through training, and then any image containing the fake information can be detected after being input into the deep neural network, so that the fake certificate image can be judged, and otherwise, the fake certificate image is a real certificate image.
In an optional mode, when the certificate image sample is a certificate image forged sample, the certificate image forged sample is also marked with forged source marking information; training the deep neural network, further comprising: acquiring counterfeiting source prediction information of a certificate image counterfeiting sample according to the comprehensive characteristics; and training the deep neural network based on the forged source prediction information and the forged source marking information.
In an alternative, training the deep neural network further includes: and when the newly added certificate image forgery sample is acquired, training the deep neural network based on the newly added certificate image forgery sample, wherein the newly added certificate image forgery sample is obtained by a newly added certificate image forgery technology.
S202: and according to the certificate counterfeit feature extraction and identification model, performing feature extraction on the certificate image to be detected.
In an alternative, the feature extraction is performed on the document image to be detected by the following steps: and correspondingly extracting comprehensive characteristic information of each category from the certificate image to be detected according to the category of the counterfeit characteristic information determined by the certificate counterfeit characteristic extraction and identification model.
For example, assume that the counterfeit feature information determined by the certificate counterfeit feature extraction and identification model includes: LBP, HSC, RGB, LARGE, SMALL and TINY, and then correspondingly extracting the six types of characteristic information from the certificate image to be detected.
S203: and carrying out counterfeiting identification on the certificate image according to the comprehensive characteristics.
In an alternative mode, certificate forgery recognition according to a plurality of categories of characteristic information is realized through the following steps: and inputting the comprehensive characteristics of a plurality of categories into a certificate counterfeit characteristic extraction and identification model, and detecting whether the certificate image to be detected contains one or more counterfeit characteristic information.
Still like the above example, six types of feature information, namely LBP, HSC, RGB, LARGE, SMALL, and TINY, are extracted from the to-be-detected document image, the six types of feature information are input into the document counterfeit feature extraction and identification model, the above feature information is identified and classified by the document counterfeit feature extraction and identification model in which the six types of counterfeit feature information have been stored, and if the extracted features are matched with one or more items of counterfeit feature information stored in the model, it is determined that the to-be-detected document image includes one or more items of counterfeit feature information.
S204: and obtaining the identification result of the certificate image.
Still as in the above example, if it is determined that the document image to be detected contains one or more counterfeit features, the output document identification result is: the certificate image is similar prompt information such as forged certificate image. In addition, because the sources of various forged certificate images are marked in the modeling process, when the certificate identification result is the forged certificate image, the sources of the forged certificate images can be simultaneously output, for example, the forged certificate image is a copy or a copy.
In an alternative, when the identification result indicates that the document image to be detected is a counterfeit document image, the method may further include the steps of: and outputting the counterfeit source of the certificate image according to the comprehensive characteristics. On this basis still include: and training the deep neural network according to the counterfeit identification result and the counterfeit source of the certificate image. Therefore, the model of the deep neural network is updated by taking the certificate image to be detected as sample data. In the prior art, the method for detecting the certificate forgery prevention is not easy to update once being determined, and is difficult to deal with the new situation of forgery. The invention carries out certificate anti-counterfeiting detection based on the deep neural network, can train the deep neural network in time through newly appeared counterfeiting information when a new counterfeiting sample appears, is easy to carry out fine tuning type training at the rear end, updates the deep neural network in time, leads the deep neural network to learn various newly appeared counterfeiting information in time and to be on-line in time, and can rapidly deal with the newly appeared counterfeiting condition for anti-counterfeiting detection.
In an optional mode, after obtaining the identification result of the certificate image, the method further comprises the following steps: if no certificate image forgery exists, extracting certificate information, and comparing the extracted certificate information with information in a certificate library; responding to the matching of the extracted certificate information and the information in the certificate library, and then requesting by a user; or responding to the fact that the extracted certificate information is not matched with the information in the certificate library, not passing a user request, marking the certificate image as a fake certificate image, and training the deep neural network.
According to the certificate identification method provided by the embodiment of the invention, when the service platform needs certificate information, certificate anti-counterfeiting detection is firstly carried out, when the video or picture to be detected passes the certificate anti-counterfeiting detection, the certificate number is extracted from the video or picture to be detected, and the image in the certificate photo is used as the identity information of the registered user. The specific process is described as follows: the service platform can acquire a video or an image of the certificate to be detected when a login or verification request is received, after frame selection operation and preprocessing operation, extracts multi-class characteristic information through a trained deep neural network, detects whether the image to be detected has counterfeit information, performs certificate anti-counterfeit detection on the image to be detected, and feeds back a certificate anti-counterfeit detection result to the service platform. If the detection result shows that no forged information exists, certificate identification and information extraction are carried out, the certificate identification and information extraction is compared with information in the library to verify the correctness of the information, if the information is correct, the login request passes, and the service request sent by the user requesting login is processed. If the detection result shows that the fake information exists, the login request does not pass, and the fake information source can be selectively returned for further distinguishing.
Therefore, in the embodiment of the invention, the modeling is performed through the strong description capability of the deep neural network, so that the discrimination of the forged characteristic information is more accurate, and the real identity information and the forged identity information can be better distinguished; and through the differentiation of the counterfeit certificate source, the counterfeit information source can be used as an additional counterfeit type label to be provided for the service port for use.
FIG. 4 is a schematic view of an embodiment of a document identification device of the present invention. The apparatus of this embodiment may be used to implement the method embodiments of the present invention described above. As shown in fig. 4, the apparatus of this embodiment includes:
an acquiring unit 401, configured to acquire a certificate image to be detected;
the processing unit 402 is configured to process a certificate image to be detected to obtain an image area to be detected including the certificate to be detected;
an extracting unit 403, configured to perform feature extraction on the to-be-detected image region according to a deep neural network to obtain a comprehensive feature;
and the identification unit 404 is configured to perform certificate image forgery identification according to the comprehensive features to obtain an identification result of the certificate image.
In one option, the integrated feature information includes: at least one or more of local binary pattern features, sparsely encoded histogram features, color features, panorama features, document region features, and document detail features.
In an optional mode, the local binary pattern feature is used for representing edge information in an image to be detected; the sparse coded histogram features are used for representing reflection and fuzzy information in an image to be detected; the color features are used for distinguishing the certificate which is really shot from the certificate copy; the panoramic image characteristics are used for extracting counterfeiting information of the image to be detected; the certificate area is used for determining a certificate area cutting chart in the image to be detected; and the certificate detail characteristics are used for extracting the counterfeiting information of the image to be detected.
FIG. 5 is a schematic view of another embodiment of a document identification device of the present invention. The apparatus of this embodiment may be used to implement the method embodiments of the present invention described above. As shown in fig. 5, the apparatus of this embodiment includes:
an acquiring unit 501, configured to acquire a certificate image to be detected;
the processing unit 502 is configured to process a certificate image to be detected to obtain an image area to be detected including the certificate to be detected;
an extracting unit 503, configured to perform feature extraction on the image region to be detected according to a deep neural network, so as to obtain a comprehensive feature;
and the identification unit 504 is used for performing certificate image forgery identification according to the comprehensive characteristics to obtain an identification result of the certificate image.
In one option, the apparatus further comprises: a training unit 505, configured to train the deep neural network, where the training unit 505 includes:
the acquisition subunit 5051 is configured to acquire each certificate image sample, where the certificate image sample is marked with information on whether to counterfeit;
an extraction subunit 5052, configured to perform feature extraction on the certificate image sample according to the deep neural network to obtain comprehensive features of the certificate image sample;
an identification subunit 5053, configured to perform certificate forgery identification according to the comprehensive feature to obtain an identification result of the certificate image sample;
a training execution subunit 5054, configured to train the deep neural network based on the recognition result of the certificate image sample and the labeled counterfeit information.
In an optional mode, when the certificate image sample is a certificate image forged sample, the certificate image forged sample is also marked with forged source marking information; the training unit 505 further comprises:
a counterfeit source acquisition sub-unit 5055, configured to acquire counterfeit source prediction information of a counterfeit sample of the certificate image according to the comprehensive feature;
a counterfeit training subunit 5056, configured to train the deep neural network based on the counterfeit source prediction information and the counterfeit source tagging information.
In an optional manner, the training unit 505 further includes:
and the new sample training subunit 5057 is configured to train the deep neural network based on the certificate image counterfeit new sample when the certificate image counterfeit new sample is acquired, and the certificate image counterfeit new sample is obtained through a certificate image new counterfeit technology.
In one alternative, the composite features include: at least one or more of local binary pattern features, sparsely encoded histogram features, color features, panorama features, document region features, and document detail features.
In an optional mode, the local binary pattern feature is used for representing edge information in an image to be detected; the sparse coded histogram features are used for representing reflection and fuzzy information in an image to be detected; the color features are used for distinguishing the certificate which is really shot from the certificate copy; the panoramic image characteristics are used for extracting counterfeiting information of the image to be detected; the certificate area is used for determining a certificate area cutting chart in the image to be detected; and the certificate detail characteristics are used for extracting the counterfeiting information of the image to be detected.
In an alternative, the identification result of the certificate image comprises: whether the certificate image is counterfeit or not; when the identification result is that the certificate image is forged, the device further comprises:
a counterfeit source output unit 506, configured to output a counterfeit source of the certificate image according to the comprehensive feature.
In one option, the apparatus further comprises:
and a result training unit 507, configured to train the deep neural network according to a counterfeit recognition result of the certificate image and the counterfeit source.
In an alternative, the document image is the document image itself or a video frame in the document video.
In an alternative, the processing unit 502 includes:
the certificate image processing subunit 5021 is used for denoising, enhancing, smoothing and sharpening the certificate image to obtain the image area to be detected; or,
and the certificate video processing subunit 5022 is configured to perform video frame selection on a certificate video to obtain a target video frame image, and perform denoising, enhancement, smoothing and sharpening on the target video frame image to obtain the to-be-detected image region.
In an optional manner, the credential video processing subunit 5022 is specifically configured to:
aiming at the video frame of the certificate video, any one or any combination of the following operations is carried out: judging whether the certificate is positioned in the center of the video frame image, whether the edge of the certificate is completely included in the video frame image, whether the area proportion of the surface of the certificate in the video frame image meets a preset proportion, whether the definition of the video frame image meets a preset definition threshold value, and whether the exposure of the image in the video meets a preset exposure threshold value;
and selecting the target video frame image from the certificate video according to the judgment result.
In an alternative, for the document video to be detected, the apparatus further comprises:
a forgery determination subunit 508, configured to determine that there is no forgery in the certificate video if there is no forgery in the identification results for all the video frame images in the certificate video; or if the identification result of any video frame image is that forgery exists, determining that the forgery exists in the certificate video.
In an alternative, after obtaining the recognition result of the document image, the apparatus further includes:
a comparing unit 509, configured to extract credential information if there is no credential image forgery, and compare the extracted credential information with information in a credential library;
a response unit 510, configured to respond to that the extracted credential information matches information in the credential library, and then request by a user; or responding to the fact that the extracted certificate information is not matched with the information in the certificate library, not requesting by a user, marking the certificate image as a forged certificate image, and training the deep neural network.
The embodiment of the invention also provides electronic equipment, which can be a mobile terminal, a Personal Computer (PC), a tablet computer, a server and the like. Referring now to fig. 6, there is shown a schematic diagram of an electronic device 600 suitable for use in implementing a terminal device or server of an embodiment of the present application: as shown in fig. 6, computer system 600 includes one or more processors, communications, etc., such as: one or more Central Processing Units (CPUs) 601, and/or one or more image processors (GPUs) 613, etc., which may perform various appropriate actions and processes according to executable instructions stored in a Read Only Memory (ROM)602 or loaded from a storage section 608 into a Random Access Memory (RAM) 603. Communications portion 612 may include, but is not limited to, a network card, which may include, but is not limited to, an ib (infiniband) network card,
the processor can communicate with the rom602 and/or the ram 630 to execute executable instructions, connect with the communication part 612 through the bus 604, and communicate with other target devices through the communication part 612, so as to complete operations corresponding to any method provided by the embodiments of the present application, for example, acquiring an image of a document to be detected; processing a certificate image to be detected to obtain an image area to be detected containing the certificate to be detected; extracting the characteristics of the image area to be detected according to the deep neural network to obtain comprehensive characteristics; and carrying out certificate image forgery identification according to the comprehensive characteristics to obtain an identification result of the certificate image.
In addition, in the RAM603, various programs and data necessary for the operation of the device can also be stored. The CPU601, ROM602, and RAM603 are connected to each other via a bus 604. The ROM602 is an optional module in case of the RAM 603. The RAM603 stores or writes executable instructions into the ROM602 at runtime, and the executable instructions cause the processor 601 to perform operations corresponding to the above-described communication method. An input/output (I/O) interface 605 is also connected to bus 604. The communication unit 612 may be integrated, or may be provided with a plurality of sub-modules (e.g., a plurality of IB network cards) and connected to the bus link.
The following components are connected to the I/O interface 605: an input portion 606 including a keyboard, a mouse, and the like; an output portion 607 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage section 608 including a hard disk and the like; and a communication section 609 including a network interface card such as a LAN card, a modem, or the like. The communication section 609 performs communication processing via a network such as the internet. The driver 610 is also connected to the I/O interface 605 as needed. A removable medium 611 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 610 as necessary, so that a computer program read out therefrom is mounted in the storage section 608 as necessary.
It should be noted that the architecture shown in fig. 6 is only an optional implementation manner, and in a specific practical process, the number and types of the components in fig. 6 may be selected, deleted, added or replaced according to actual needs; in different functional component settings, separate settings or integrated settings may also be used, for example, the GPU and the CPU may be separately set or the GPU may be integrated on the CPU, the communication part may be separately set or integrated on the CPU or the GPU, and so on. These alternative embodiments are all within the scope of the present disclosure.
In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program tangibly embodied on a machine-readable medium, the computer program comprising program code for performing the method illustrated in the flow chart, the program code may include instructions corresponding to performing the method steps provided by embodiments of the present application, e.g., receiving a picture or video of a document to be detected; processing the picture or video of the certificate to be detected to obtain a certificate image to be detected; extracting features of a certificate image to be detected to obtain feature information of a plurality of categories; and carrying out certificate forgery identification according to the characteristic information of the multiple categories to obtain the identification result of the certificate. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 609, and/or installed from the removable medium 611. The computer program performs the above-described functions defined in the method of the present application when executed by a Central Processing Unit (CPU) 601.
Those of ordinary skill in the art will understand that: all or part of the steps for implementing the method embodiments may be implemented by hardware related to program instructions, and the program may be stored in a computer readable storage medium, and when executed, the program performs the steps including the method embodiments; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
The method and apparatus of the present invention may be implemented in a number of ways. For example, the methods and apparatus of the present invention may be implemented in software, hardware, firmware, or any combination of software, hardware, and firmware. The above-described order for the steps of the method is for illustrative purposes only, and the steps of the method of the present invention are not limited to the order specifically described above unless specifically indicated otherwise. Furthermore, in some embodiments, the present invention may also be embodied as a program recorded in a recording medium, the program including machine-readable instructions for implementing a method according to the present invention. Thus, the present invention also covers a recording medium storing a program for executing the method according to the present invention. The description of the present invention has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the invention in the form disclosed. Many modifications and variations will be apparent to practitioners skilled in this art. The embodiment was chosen and described in order to best explain the principles of the invention and the practical application, and to enable others of ordinary skill in the art to understand the invention for various embodiments with various modifications as are suited to the particular use contemplated.
Claims (10)
1. A method of document identification, comprising:
acquiring a certificate image to be detected;
processing a certificate image to be detected to obtain an image area to be detected containing the certificate to be detected;
extracting the characteristics of the image area to be detected according to the deep neural network to obtain comprehensive characteristics;
and carrying out certificate image forgery identification according to the comprehensive characteristics to obtain an identification result of the certificate image.
2. The method of claim 1, further comprising:
training the deep neural network;
the training the deep neural network comprises:
collecting each certificate image sample, wherein the certificate image sample is marked with information whether to be forged or not;
extracting the features of the certificate image sample according to the deep neural network to obtain the comprehensive features of the certificate image sample;
carrying out certificate forgery identification according to the comprehensive characteristics to obtain an identification result of the certificate image sample;
and training the deep neural network based on the identification result of the certificate image sample and the marked fake information.
3. The method according to claim 2, wherein when the certificate image sample is a certificate image counterfeit sample, the certificate image counterfeit sample is further marked with counterfeit source marking information; the training the deep neural network further comprises:
acquiring counterfeiting source prediction information of a certificate image counterfeiting sample according to the comprehensive characteristics;
and training the deep neural network based on the forged source prediction information and the forged source marking information.
4. The method of claim 2, wherein the training the deep neural network further comprises:
and when a certificate image forgery new sample is acquired, training the deep neural network based on the certificate image forgery new sample, wherein the certificate image forgery new sample is obtained by a certificate image forgery technology.
5. The method according to any of claims 2-4, wherein the comprehensive characteristics comprise: at least one or more of local binary pattern features, sparsely encoded histogram features, color features, panorama features, document region features, and document detail features.
6. The method according to claim 5, wherein the local binary pattern features are used for characterizing edge information in an image to be detected; the sparse coded histogram features are used for representing reflection and fuzzy information in an image to be detected; the color features are used for distinguishing the certificate which is really shot from the certificate copy; the panoramic image characteristics are used for extracting counterfeiting information of the image to be detected; the certificate area is used for determining a certificate area cutting chart in the image to be detected; and the certificate detail characteristics are used for extracting the counterfeiting information of the image to be detected.
7. The method according to any one of claims 1-4, wherein the recognition result of the document image comprises: whether the certificate image is counterfeit or not; when the identification result is that the certificate image is forged, the method further comprises the following steps:
and outputting a counterfeit source of the certificate image according to the comprehensive characteristics.
8. A document identification device, comprising:
the acquisition unit is used for acquiring a certificate image to be detected;
the processing unit is used for processing a certificate image to be detected to obtain an image area to be detected containing the certificate to be detected;
the extraction unit is used for extracting the characteristics of the image area to be detected according to the deep neural network to obtain comprehensive characteristics;
and the identification unit is used for carrying out certificate image forgery identification according to the comprehensive characteristics to obtain the identification result of the certificate image.
9. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the steps of the method of any of claims 1-7 are implemented when the program is executed by the processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711052744.8A CN108229499A (en) | 2017-10-30 | 2017-10-30 | Certificate recognition methods and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711052744.8A CN108229499A (en) | 2017-10-30 | 2017-10-30 | Certificate recognition methods and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108229499A true CN108229499A (en) | 2018-06-29 |
Family
ID=62655722
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711052744.8A Pending CN108229499A (en) | 2017-10-30 | 2017-10-30 | Certificate recognition methods and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108229499A (en) |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109255299A (en) * | 2018-08-13 | 2019-01-22 | 北京市商汤科技开发有限公司 | Identity identifying method and device, electronic equipment and storage medium |
| CN109359502A (en) * | 2018-08-13 | 2019-02-19 | 北京市商汤科技开发有限公司 | False-proof detection method and device, electronic equipment, storage medium |
| CN109815960A (en) * | 2018-12-21 | 2019-05-28 | 深圳壹账通智能科技有限公司 | Reproduction image-recognizing method, device, equipment and medium based on deep learning |
| CN110070130A (en) * | 2019-04-23 | 2019-07-30 | 杭州智趣智能信息技术有限公司 | A kind of certificate information recognition methods, transposition, equipment and medium |
| CN110309859A (en) * | 2019-06-06 | 2019-10-08 | 阿里巴巴集团控股有限公司 | A kind of image true-false detection method, device and electronic equipment |
| CN110443237A (en) * | 2019-08-06 | 2019-11-12 | 北京旷视科技有限公司 | Certificate recognition methods, device, electronic equipment and computer readable storage medium |
| CN110533643A (en) * | 2019-08-21 | 2019-12-03 | 阿里巴巴集团控股有限公司 | Certificate identification method and device |
| CN110717450A (en) * | 2019-10-09 | 2020-01-21 | 深圳大学 | Training method and detection method for automatically identifying copied image of original document |
| WO2020034733A1 (en) * | 2018-08-13 | 2020-02-20 | 北京市商汤科技开发有限公司 | Identity authentication method and apparatus, electronic device, and storage medium |
| CN111127398A (en) * | 2019-11-25 | 2020-05-08 | 深圳前海微众银行股份有限公司 | Method and device for detecting certificate photo duplication, terminal equipment and storage medium |
| CN111259894A (en) * | 2020-01-20 | 2020-06-09 | 普信恒业科技发展（北京）有限公司 | Certificate information identification method and device and computer equipment |
| CN111275042A (en) * | 2020-01-21 | 2020-06-12 | 支付宝实验室(新加坡)有限公司 | Counterfeit certificate identification method and device and electronic equipment |
| CN111291586A (en) * | 2018-12-06 | 2020-06-16 | 北京市商汤科技开发有限公司 | Living body detection method, living body detection device, electronic apparatus, and computer-readable storage medium |
| CN111324874A (en) * | 2020-01-21 | 2020-06-23 | 支付宝实验室(新加坡)有限公司 | Certificate authenticity identification method and device |
| CN111353434A (en) * | 2020-02-28 | 2020-06-30 | 北京市商汤科技开发有限公司 | Information identification method, device, system, electronic equipment and storage medium |
| US10769263B1 (en) | 2019-05-07 | 2020-09-08 | Alibaba Group Holding Limited | Certificate verification |
| CN111652223A (en) * | 2019-08-21 | 2020-09-11 | 阿里巴巴集团控股有限公司 | Certificate identification method and device |
| CN111767828A (en) * | 2020-06-28 | 2020-10-13 | 京东数字科技控股有限公司 | Certificate image copying and identifying method and device, electronic equipment and storage medium |
| CN111898520A (en) * | 2020-07-28 | 2020-11-06 | 腾讯科技（深圳）有限公司 | Certificate authenticity identification method and device, computer readable medium and electronic equipment |
| WO2020224296A1 (en) * | 2019-05-07 | 2020-11-12 | 创新先进技术有限公司 | Certificate verification and identity verification methods and apparatuses, and devices |
| CN112465807A (en) * | 2020-12-14 | 2021-03-09 | 深圳市芊熠智能硬件有限公司 | License plate image authenticity identification method, device, equipment and medium |
| CN112528998A (en) * | 2021-02-18 | 2021-03-19 | 成都新希望金融信息有限公司 | Certificate image processing method and device, electronic equipment and readable storage medium |
| CN112613497A (en) * | 2020-02-21 | 2021-04-06 | 支付宝实验室(新加坡)有限公司 | Method and system for determining the authenticity of official documents |
| CN113590286A (en) * | 2021-07-28 | 2021-11-02 | 北京市商汤科技开发有限公司 | Task processing system and task processing method |
| CN114299371A (en) * | 2021-11-30 | 2022-04-08 | 深圳壹账通智能科技有限公司 | Method, system, device and medium for certificate recognition model training and certificate recognition |
| WO2022089124A1 (en) * | 2020-10-29 | 2022-05-05 | 腾讯科技（深圳）有限公司 | Certificate authenticity identification method and apparatus, computer-readable medium, and electronic device |
| CN114743016A (en) * | 2022-04-15 | 2022-07-12 | 成都新希望金融信息有限公司 | Certificate authenticity identification method and device, electronic equipment and storage medium |
| CN114819142A (en) * | 2022-04-18 | 2022-07-29 | 支付宝(杭州)信息技术有限公司 | Screen image recognition method and device, training method and device of model of screen image recognition method and device, and electronic equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102156853A (en) * | 2010-02-11 | 2011-08-17 | 深圳鼎识科技有限公司 | Certificate reading system and certificate reading method |
| CN104079587A (en) * | 2014-07-21 | 2014-10-01 | 深圳天祥质量技术服务有限公司 | Certificate identification device and certificate check system |
| CN105118048A (en) * | 2015-07-17 | 2015-12-02 | 北京旷视科技有限公司 | Method and device for identifying copying certificate image |
| US20170078281A1 (en) * | 2015-09-15 | 2017-03-16 | Solid Year Co., Ltd. | User Login Method and System Capable of Analyzing User Face Validity |
| CN106778607A (en) * | 2016-12-15 | 2017-05-31 | 国政通科技股份有限公司 | A kind of people based on recognition of face and identity card homogeneity authentication device and method |
| CN106991451A (en) * | 2017-04-14 | 2017-07-28 | 武汉神目信息技术有限公司 | A kind of identifying system and method for certificate picture |
| CN106991390A (en) * | 2017-03-30 | 2017-07-28 | 电子科技大学 | A kind of hand-held testimony of a witness Compare System and method based on deep learning |
-
2017
- 2017-10-30 CN CN201711052744.8A patent/CN108229499A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102156853A (en) * | 2010-02-11 | 2011-08-17 | 深圳鼎识科技有限公司 | Certificate reading system and certificate reading method |
| CN104079587A (en) * | 2014-07-21 | 2014-10-01 | 深圳天祥质量技术服务有限公司 | Certificate identification device and certificate check system |
| CN105118048A (en) * | 2015-07-17 | 2015-12-02 | 北京旷视科技有限公司 | Method and device for identifying copying certificate image |
| US20170078281A1 (en) * | 2015-09-15 | 2017-03-16 | Solid Year Co., Ltd. | User Login Method and System Capable of Analyzing User Face Validity |
| CN106778607A (en) * | 2016-12-15 | 2017-05-31 | 国政通科技股份有限公司 | A kind of people based on recognition of face and identity card homogeneity authentication device and method |
| CN106991390A (en) * | 2017-03-30 | 2017-07-28 | 电子科技大学 | A kind of hand-held testimony of a witness Compare System and method based on deep learning |
| CN106991451A (en) * | 2017-04-14 | 2017-07-28 | 武汉神目信息技术有限公司 | A kind of identifying system and method for certificate picture |
Non-Patent Citations (2)
| Title |
|---|
| 张宜浩、文俊浩: "《基于半监督学习的个性化推荐算法研究》", 31 May 2016, 重庆大学出版社 * |
| 葛晓滨 等: "《移动电子商务教程》", 30 September 2014, 中国科学技术大学出版社 * |
Cited By (47)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102406432B1 (en) * | 2018-08-13 | 2022-06-08 | 베이징 센스타임 테크놀로지 디벨롭먼트 컴퍼니 리미티드 | Identity authentication methods and devices, electronic devices and storage media |
| CN109359502A (en) * | 2018-08-13 | 2019-02-19 | 北京市商汤科技开发有限公司 | False-proof detection method and device, electronic equipment, storage medium |
| KR20200118842A (en) * | 2018-08-13 | 2020-10-16 | 베이징 센스타임 테크놀로지 디벨롭먼트 컴퍼니 리미티드 | Identity authentication method and device, electronic device and storage medium |
| JP2021516819A (en) * | 2018-08-13 | 2021-07-08 | ベイジン センスタイム テクノロジー デベロップメント カンパニー， リミテッド | ID authentication method and devices, electronic devices and storage media |
| CN109255299A (en) * | 2018-08-13 | 2019-01-22 | 北京市商汤科技开发有限公司 | Identity identifying method and device, electronic equipment and storage medium |
| JP7165746B2 (en) | 2018-08-13 | 2022-11-04 | ベイジン・センスタイム・テクノロジー・デベロップメント・カンパニー・リミテッド | ID authentication method and device, electronic device and storage medium |
| WO2020034733A1 (en) * | 2018-08-13 | 2020-02-20 | 北京市商汤科技开发有限公司 | Identity authentication method and apparatus, electronic device, and storage medium |
| CN111291586A (en) * | 2018-12-06 | 2020-06-16 | 北京市商汤科技开发有限公司 | Living body detection method, living body detection device, electronic apparatus, and computer-readable storage medium |
| CN111291586B (en) * | 2018-12-06 | 2024-05-24 | 北京市商汤科技开发有限公司 | Living body detection method, living body detection device, electronic equipment and computer readable storage medium |
| CN109815960A (en) * | 2018-12-21 | 2019-05-28 | 深圳壹账通智能科技有限公司 | Reproduction image-recognizing method, device, equipment and medium based on deep learning |
| CN110070130A (en) * | 2019-04-23 | 2019-07-30 | 杭州智趣智能信息技术有限公司 | A kind of certificate information recognition methods, transposition, equipment and medium |
| US11392679B2 (en) | 2019-05-07 | 2022-07-19 | Advanced New Technologies Co., Ltd. | Certificate verification |
| US10909227B2 (en) | 2019-05-07 | 2021-02-02 | Advanced New Technologies Co., Ltd. | Certificate verification |
| WO2020224296A1 (en) * | 2019-05-07 | 2020-11-12 | 创新先进技术有限公司 | Certificate verification and identity verification methods and apparatuses, and devices |
| US11086977B2 (en) | 2019-05-07 | 2021-08-10 | Advanced New Technologies Co., Ltd. | Certificate verification |
| US10769263B1 (en) | 2019-05-07 | 2020-09-08 | Alibaba Group Holding Limited | Certificate verification |
| CN110309859B (en) * | 2019-06-06 | 2023-04-07 | 创新先进技术有限公司 | Image authenticity detection method and device and electronic equipment |
| CN110309859A (en) * | 2019-06-06 | 2019-10-08 | 阿里巴巴集团控股有限公司 | A kind of image true-false detection method, device and electronic equipment |
| CN110443237B (en) * | 2019-08-06 | 2023-06-30 | 北京旷视科技有限公司 | Certificate identification method, device, electronic equipment and computer readable storage medium |
| CN110443237A (en) * | 2019-08-06 | 2019-11-12 | 北京旷视科技有限公司 | Certificate recognition methods, device, electronic equipment and computer readable storage medium |
| CN111652223A (en) * | 2019-08-21 | 2020-09-11 | 阿里巴巴集团控股有限公司 | Certificate identification method and device |
| CN110533643A (en) * | 2019-08-21 | 2019-12-03 | 阿里巴巴集团控股有限公司 | Certificate identification method and device |
| CN110533643B (en) * | 2019-08-21 | 2023-06-30 | 创新先进技术有限公司 | Certificate authentication method and device |
| CN111652223B (en) * | 2019-08-21 | 2023-04-18 | 创新先进技术有限公司 | Certificate identification method and device |
| CN110717450A (en) * | 2019-10-09 | 2020-01-21 | 深圳大学 | Training method and detection method for automatically identifying copied image of original document |
| CN111127398B (en) * | 2019-11-25 | 2023-09-26 | 深圳前海微众银行股份有限公司 | Certificate photo copying detection method and device, terminal equipment and storage medium |
| CN111127398A (en) * | 2019-11-25 | 2020-05-08 | 深圳前海微众银行股份有限公司 | Method and device for detecting certificate photo duplication, terminal equipment and storage medium |
| CN111259894B (en) * | 2020-01-20 | 2023-07-07 | 普信恒业科技发展（北京）有限公司 | Certificate information identification method and device and computer equipment |
| CN111259894A (en) * | 2020-01-20 | 2020-06-09 | 普信恒业科技发展（北京）有限公司 | Certificate information identification method and device and computer equipment |
| CN111275042A (en) * | 2020-01-21 | 2020-06-12 | 支付宝实验室(新加坡)有限公司 | Counterfeit certificate identification method and device and electronic equipment |
| CN111324874A (en) * | 2020-01-21 | 2020-06-23 | 支付宝实验室(新加坡)有限公司 | Certificate authenticity identification method and device |
| CN111275042B (en) * | 2020-01-21 | 2023-07-18 | 支付宝实验室(新加坡)有限公司 | Identification method and device for forged certificate and electronic equipment |
| CN112613497A (en) * | 2020-02-21 | 2021-04-06 | 支付宝实验室(新加坡)有限公司 | Method and system for determining the authenticity of official documents |
| WO2021169384A1 (en) * | 2020-02-28 | 2021-09-02 | 北京市商汤科技开发有限公司 | Information identification method and apparatus, and system, electronic device, storage medium and computer program |
| JP2022524672A (en) * | 2020-02-28 | 2022-05-10 | 北京市商▲湯▼科技▲開▼▲發▼有限公司 | Information recognition methods and devices, systems, electronic devices, storage media and computer programs |
| TWI766458B (en) * | 2020-02-28 | 2022-06-01 | 大陸商北京市商湯科技開發有限公司 | Information identification method and apparatus, electronic device, and storage medium |
| CN111353434A (en) * | 2020-02-28 | 2020-06-30 | 北京市商汤科技开发有限公司 | Information identification method, device, system, electronic equipment and storage medium |
| CN111767828A (en) * | 2020-06-28 | 2020-10-13 | 京东数字科技控股有限公司 | Certificate image copying and identifying method and device, electronic equipment and storage medium |
| CN111767828B (en) * | 2020-06-28 | 2023-12-08 | 京东科技控股股份有限公司 | Certificate image reproduction identification method and device, electronic equipment and storage medium |
| CN111898520A (en) * | 2020-07-28 | 2020-11-06 | 腾讯科技（深圳）有限公司 | Certificate authenticity identification method and device, computer readable medium and electronic equipment |
| WO2022089124A1 (en) * | 2020-10-29 | 2022-05-05 | 腾讯科技（深圳）有限公司 | Certificate authenticity identification method and apparatus, computer-readable medium, and electronic device |
| CN112465807A (en) * | 2020-12-14 | 2021-03-09 | 深圳市芊熠智能硬件有限公司 | License plate image authenticity identification method, device, equipment and medium |
| CN112528998A (en) * | 2021-02-18 | 2021-03-19 | 成都新希望金融信息有限公司 | Certificate image processing method and device, electronic equipment and readable storage medium |
| CN113590286A (en) * | 2021-07-28 | 2021-11-02 | 北京市商汤科技开发有限公司 | Task processing system and task processing method |
| CN114299371A (en) * | 2021-11-30 | 2022-04-08 | 深圳壹账通智能科技有限公司 | Method, system, device and medium for certificate recognition model training and certificate recognition |
| CN114743016A (en) * | 2022-04-15 | 2022-07-12 | 成都新希望金融信息有限公司 | Certificate authenticity identification method and device, electronic equipment and storage medium |
| CN114819142A (en) * | 2022-04-18 | 2022-07-29 | 支付宝(杭州)信息技术有限公司 | Screen image recognition method and device, training method and device of model of screen image recognition method and device, and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108229499A (en) | Certificate recognition methods and device, electronic equipment and storage medium | |
| Scherhag et al. | Detection of face morphing attacks based on PRNU analysis | |
| US11087125B2 (en) | Document authenticity determination | |
| US20220406029A1 (en) | Tamper detection for identification documents | |
| Mahmood et al. | Copy‐move forgery detection technique for forensic analysis in digital images | |
| US20200005035A1 (en) | Range and/or polarity-based thresholding for improved data extraction | |
| KR102406432B1 (en) | Identity authentication methods and devices, electronic devices and storage media | |
| US9152860B2 (en) | Methods and apparatus for capturing, processing, training, and detecting patterns using pattern recognition classifiers | |
| US11023708B2 (en) | Within document face verification | |
| CN110569878A (en) | Photograph background similarity clustering method based on convolutional neural network and computer | |
| CN109255299A (en) | Identity identifying method and device, electronic equipment and storage medium | |
| WO2016131083A1 (en) | Identity verification. method and system for online users | |
| CN109271915B (en) | Anti-counterfeiting detection method and device, electronic equipment and storage medium | |
| US10043071B1 (en) | Automated document classification | |
| Ding et al. | A novel method for detecting image sharpening based on local binary pattern | |
| CN112257709A (en) | Signboard photo auditing method and device, electronic equipment and readable storage medium | |
| Sirajudeen et al. | Forgery document detection in information management system using cognitive techniques | |
| Nowroozi et al. | Detecting high-quality GAN-generated face images using neural networks | |
| CN113111880A (en) | Certificate image correction method and device, electronic equipment and storage medium | |
| CN111259894B (en) | Certificate information identification method and device and computer equipment | |
| CN113077355B (en) | Insurance claim settlement method and device, electronic equipment and storage medium | |
| CN110415424B (en) | Anti-counterfeiting identification method and device, computer equipment and storage medium | |
| CN115359302A (en) | Coin identification method, system and storage medium | |
| Markham et al. | Open-Set: ID Card Presentation Attack Detection Using Neural Style Transfer | |
| Tapia et al. | First Competition on Presentation Attack Detection on ID Card |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180629 |