CN107481238A - Image quality measure method and device - Google Patents
Image quality measure method and device Download PDFInfo
- Publication number
- CN107481238A CN107481238A CN201710854415.9A CN201710854415A CN107481238A CN 107481238 A CN107481238 A CN 107481238A CN 201710854415 A CN201710854415 A CN 201710854415A CN 107481238 A CN107481238 A CN 107481238A
- Authority
- CN
- China
- Prior art keywords
- target area
- image
- quality
- profile
- quality evaluation
- 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 46
- 238000013441 quality evaluation Methods 0.000 claims abstract description 83
- 238000001303 quality assessment method Methods 0.000 claims abstract description 56
- 238000001514 detection method Methods 0.000 claims abstract description 32
- 230000003628 erosive effect Effects 0.000 claims description 9
- 238000005530 etching Methods 0.000 claims description 7
- 238000011156 evaluation Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- 238000012549 training Methods 0.000 description 6
- 238000013135 deep learning Methods 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000005260 corrosion Methods 0.000 description 2
- 230000007797 corrosion Effects 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 241001465754 Metazoa Species 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000003139 buffering effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 230000003245 working effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- 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
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- 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
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
-
- 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
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20092—Interactive image processing based on input by user
- G06T2207/20104—Interactive definition of region of interest [ROI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)
Abstract
One aspect of the present invention provides a kind of image quality measure method and device.The image quality measure method includes：Target detection is carried out to image to be assessed, to determine at least one target area；Quality evaluation is carried out respectively at least one target area, to determine the quality assessment result of each target area at least one target area；Based on the quality assessment result of each target area in identified at least one target area, quality evaluation is carried out to image to be assessed.Image quality measure method provided by the present invention causes image quality measure process to pay close attention to target area interested, ignore the picture quality in inessential region, quality evaluation is carried out to image to be assessed so as to realize, estimating velocity is fast, Evaluation accuracy is higher, the objective quality that effectively have evaluated image to be assessed.
Description
Technical field
The present invention relates to digital image processing techniques field, more particularly to a kind of image quality measure method and device.
Background technology
With the popularization of the digital equipments such as digital camera, monitoring camera, mobile phone, digital picture is using more and more extensive.
And in practical application, such as：Recognition of face, OCR, image classification, intelligent monitoring etc., also there is certain want to the quality of image
Ask, how effectively to assess the quality of a sub-picture quality has become more and more important.Current image quality measure method,
Either there is the mode of the reference still assessment mode without reference, be all substantially based on and whole sub-picture is handled, obtain
Its corresponding quality index.Regarded in fact, not embodying human eye when computer carries out quality evaluation in this way really
The processing procedure of feel.Human visual system can realize auto-focusing when assessing surrounding scene, and this focus process is also with regard to quite
In the real target area interested of searching, that is, target detection.The quality of area-of-interest (ROI) is answered actual in image
It is more meaningful with.
Therefore, a kind of image quality measure method and apparatus are needed badly so that image quality measure process is more paid close attention to interested
Target area, ignore the picture quality in inessential region.
The content of the invention
The present invention is in view of the above-mentioned problems, propose a kind of image quality measure method and device.
One aspect of the present invention provides a kind of image quality measure method, and methods described includes：Image to be assessed is carried out
Target detection, to determine at least one target area；Quality evaluation is carried out respectively at least one target area, to determine
The quality assessment result of each target area at least one target area；Based on identified at least one mesh
The quality assessment result of each target area in region is marked, quality evaluation is carried out to the image to be assessed.
In one embodiment, the step of carrying out target detection to the image to be assessed includes：To from described to be evaluated
The profile for estimating the binary image that image obtains is classified, to determine text profile；Based on identified text profile, referring to
Determine to merge text on direction, to determine at least one target area.
In one embodiment, based on determined by following at least one operation parameter come to the binary image
Profile is classified：Determine the non-zero pixels number in the profile of the binary image；Determine the wheel of the binary image
Wide depth-width ratio and the ratio of width to height；Determine the assigned direction neighborhood model in the range of the assigned direction of the profile of the binary image
Enclose the number of interior existing similar width profile and the number of similar height profile.
In one embodiment, based on identified text profile, the step merged in the direction indicated to text
Suddenly include：By setting Expanded Operators and erosion operator, expansive working is carried out to identified text profile in the direction indicated
And etching operation.
In one embodiment, the step of carrying out quality evaluation respectively at least one target area includes：Base
In identified at least one target area and image quality measure model, determine at least one target area
The quality assessment result of each target area.
In one embodiment, the step of carrying out quality evaluation respectively at least one target area includes：Base
Come to carry out quality evaluation to each target area at least one target area in the statistics to grey scale pixel value.
In one embodiment, based on the matter of each target area in identified at least one target area
The step of measuring assessment result, quality evaluation is carried out to the image to be assessed includes：By way of mass-weighted average, it is based on
The quality assessment result of each target area in identified at least one target area, enters to the image to be assessed
Row quality evaluation.
In one embodiment, quality evaluation is carried out respectively at least one target area, with determine described in extremely
The step of quality assessment result of each target area in a few target area, includes：Referred to using reference-free quality evaluation
Mark, quality evaluation is carried out at least one target area respectively, each at least one target area to determine
The quality assessment result of target area, wherein, the reference-free quality evaluation index includes edge strength, noise rate or unified bright
It is at least one in degree distribution.
Another aspect of the present invention provides a kind of image quality measure device, and described device includes：Object detection unit, quilt
It is configured to carry out target detection to image to be assessed, to determine at least one target area；Target area quality estimation unit, quilt
It is configured to carry out quality evaluation respectively at least one target area, it is every at least one target area to determine
The quality assessment result of individual target area；Image quality measure unit to be assessed, it is configured as described in based on determined by least
The quality assessment result of each target area in one target area, quality evaluation is carried out to the image to be assessed.
In one embodiment, the object detection unit includes：Profile taxon, it is configured as from described to be evaluated
The profile for estimating the binary image that image obtains is classified, to determine text profile；Text combining unit, it is configured as being based on
Identified text profile, is merged to text in the direction indicated, to determine at least one target area.
In one embodiment, the profile taxon is additionally configured to：Based on following at least one operation institute really
Fixed parameter is classified to the profile of the binary image：Determine the non-zero pixels in the profile of the binary image
Number；Determine the depth-width ratio and the ratio of width to height of the profile of the binary image；And determine the profile of the binary image
Assigned direction in the range of assigned direction contiguous range in existing similar width profile number and similar height profile
Number.
In one embodiment, the text combining unit is additionally configured to：By setting Expanded Operators and corrosion to calculate
Son, expansive working and etching operation are carried out to identified text profile in the direction indicated.
In one embodiment, the target area quality estimation unit is additionally configured to：Based on identified described
At least one target area and image quality measure model, determine each target area at least one target area
Quality assessment result.
In one embodiment, the target area quality estimation unit is additionally configured to：Based on to grey scale pixel value
Statistics come at least one target area each target area carry out quality evaluation.
In one embodiment, the image quality measure unit to be assessed is additionally configured to：It is flat by quality weighting
Equal mode, based on the quality assessment result of each target area in identified at least one target area, to institute
State image to be assessed and carry out quality evaluation.
In one embodiment, the target area quality estimation unit is additionally configured to：Commented using without reference mass
Valency index, quality evaluation is carried out respectively at least one target area, to determine at least one target area
The quality assessment result of each target area, wherein, the reference-free quality evaluation index includes edge strength, noise rate or system
It is at least one in one Luminance Distribution.
Another aspect of the present invention additionally provides a kind of computer-readable storage medium, is stored thereon with processor and can perform journey
Sequence, when executable program described in the computing device, follow the steps below：Target detection is carried out to image to be assessed, with
Determine at least one target area；Quality evaluation is carried out respectively at least one target area, with least one described in determination
The quality assessment result of each target area in individual target area；Based in identified at least one target area
The quality assessment result of each target area, quality evaluation is carried out to the image to be assessed.
In one embodiment, when executable program described in the computing device, the image to be assessed is entered
The step of row target detection, includes：The profile of binary image to being obtained from the image to be assessed is classified, to determine
Text profile；Based on identified text profile, text is merged in the direction indicated, to determine at least one mesh
Mark region.
In one embodiment, when executable program described in the computing device, based on following at least one behaviour
Parameter determined by work is classified to the profile of the binary image：Determine non-in the profile of the binary image
Zero number of pixels；Determine the depth-width ratio and the ratio of width to height of the profile of the binary image；Determine the wheel of the binary image
The number and similar height profile of existing similar width profile in assigned direction contiguous range in the range of wide assigned direction
Number.
In one embodiment, when executable program described in the computing device, based on identified text wheel
Exterior feature, the step of being merged in the direction indicated to text, include：By setting Expanded Operators and erosion operator, in assigned direction
On expansive working and etching operation are carried out to identified text profile.
In one embodiment, when executable program described in the computing device, at least one target
Region carries out the step of quality evaluation respectively to be included：Based on identified at least one target area and image quality measure
Model, determine the quality assessment result of each target area at least one target area.
In one embodiment, when executable program described in the computing device, at least one target
Region carries out the step of quality evaluation respectively to be included：Based on the statistics to grey scale pixel value come at least one target area
In each target area carry out quality evaluation.
In one embodiment, when executable program described in the computing device, based on it is identified it is described extremely
The quality assessment result of each target area in a few target area, the step of quality evaluation is carried out to the image to be assessed
Suddenly include：By way of mass-weighted average, based on each target area in identified at least one target area
The quality assessment result in domain, quality evaluation is carried out to the image to be assessed.
In one embodiment, when executable program described in the computing device, at least one target
Region carries out quality evaluation respectively, to determine the quality assessment result of each target area at least one target area
The step of include：Using reference-free quality evaluation index, quality evaluation is carried out respectively at least one target area, with true
The quality assessment result of each target area in fixed at least one target area, wherein, the reference-free quality evaluation
Index includes at least one in edge strength, noise rate or uniform brightness distribution.
Image quality measure method provided by the present invention causes image quality measure process by paying close attention to mesh interested
Region is marked, ignores the picture quality in inessential region, quality evaluation, estimating velocity are carried out to image to be assessed so as to realize
Hurry up, Evaluation accuracy is higher, the objective quality that effectively have evaluated image to be assessed, make quality assessment result as far as possible with human eye sense
Know and be consistent.
Brief description of the drawings
Fig. 1 is the flow chart of image quality measure method according to embodiments of the present invention；
Fig. 2 is the flow chart of text image method for evaluating quality according to embodiments of the present invention；
Fig. 3 is the example of a text image according to embodiments of the present invention；
Fig. 4 is the binary image of Fig. 3 text image；
Fig. 5 is the schematic diagram in the line of text region to being obtained after Fig. 3 progress target detections；
Fig. 6 is a schematic diagram of the composograph of image quality measure model according to embodiments of the present invention；
Fig. 7 is the assessment result of Fig. 3 text image；
Fig. 8 is the example of another text image according to embodiments of the present invention；
Fig. 9 is the binary image of Fig. 8 text image；
Figure 10 is the schematic diagram in line of text region that Fig. 8 obtained after target detection；
Figure 11 is the assessment result of Fig. 8 text image；
Figure 12 is the schematic diagram of image quality measure device according to embodiments of the present invention.
Embodiment
In the specific descriptions of following preferred embodiment, by with reference to the appended accompanying drawing for forming a present invention part.Institute
Attached accompanying drawing, which has been illustrated by way of example, can realize specific embodiment.The embodiment of example is not intended as
Limit is according to all embodiments of the invention.It is appreciated that without departing from the scope of the present invention, other can be utilized
Embodiment, structural or logicality modification can also be carried out.Therefore, following specific descriptions and nonrestrictive, and this
The scope of invention is defined by the claims appended hereto.
Fig. 1 is the flow chart of image quality measure method according to embodiments of the present invention.
The invention provides a kind of method for evaluating quality, this method includes step as shown in Figure 1：
S101：Target detection is carried out to image to be assessed, to determine at least one target area；
S102：Quality evaluation is carried out respectively at least one target area, to determine at least one target area
Each target area quality assessment result；
S103：Based on the quality assessment result of each target area of identified at least one target area, treat
Assess image and carry out quality evaluation.
In order to assess picture quality, it is necessary first to the target area in detection image.It should be understood that target interested
Region is different, different using object detection method.Image to be assessed can be used and carried out based on traditional feature extraction mode
Target detection, the mode based on deep neural network carry out target detection (such as CTPN methods), the target detection side based on statistics
Method (such as Haar classifier), object detection method (such as Fater-rcnn) based on deep learning etc., wherein, it is to be assessed
Image can include any object, for example, animal, face, food, automobile or line of text image etc..This paper image to be assessed
Can be coloured image or gray level image.The result returned after target detection is multiple targets interested
Region, these target areas preserve in the form of images.
Herein, quality evaluation uses reference-free quality evaluation index, specifically includes edge strength, noise rate or unified bright
Degree distribution etc..In addition, the method training quality evaluation model based on machine learning can be used before quality evaluation, for
Quality evaluation is carried out to the image to be assessed of input.Model training needs to collect various target area images in advance, and carries out matter
Amount mark, the size of quality annotation, for example by the integer representation in the range of numerical value 0-100, mark the bigger expression corresponding diagram of numerical value
As quality is better, it should be appreciated that other suitable modes can also be used to be defined the size of quality annotation.Implement in one kind
In mode, model training uses the training method based on deep learning (convolutional neural networks CNN), to determine high accuracy
Picture quality model.
In addition, herein, the quality assessment result of each target area at least one target area based on determination
When carrying out quality evaluation to image to be assessed, commented using the quality to each aimed quality region at least one target area
Estimate result and be weighted average mode, weights, the weight of target area can be assigned respectively according to the importance of each target area
The property wanted depends on the size of target area, definition, to interest level of the target area etc..When thinking that target area has
During equal importance, average weighted mode develops into the arithmetic average of the quality assessment result of each target area.
Make carry out further detailed description to embodiment below in conjunction with accompanying drawing.
Fig. 2 is according to the flow chart of the text image method for evaluating quality of the present invention, as shown in Fig. 2 this method is included such as
Lower step：
S201：Input image to be assessed；
In this step, image to be assessed is text image (such as shown in Fig. 3 or Fig. 8), and line of text is to feel emerging in image
The target of interest, the purpose for carrying out target detection to line of text is to detect the position of line of text in the picture.
S202：If image to be assessed is coloured image, step S103 is performed, S103 is otherwise skipped and performs step
S204；
S203：Gray processing processing is carried out to coloured image；
S204：Binary conversion treatment is carried out to pending gray level image；
In this step, when the image to be assessed of input is gray level image (such as shown in Fig. 3 and Fig. 8), figure to be assessed
As pending gray level image, without step S203 after execution of step S202, step S204 is directly arrived, to be assessed
Image carries out binary conversion treatment；When the image to be assessed of input is coloured image, gray level image to be assessed is by step
The image of S203 gray processings processing.Two are carried out to pending gray level image using local auto-adaptive binarization method in the present embodiment
Value is handled, i.e., the binary-state threshold on location of pixels is determined according to the pixel Distribution value of neighborhood of pixels, to pending gray scale
All pixels point in image, perform following operation：
Centered on pixel, N × N neighborhoods region is chosen, in one embodiment, N is between [2,5]；
Calculate the average of all pixels in the contiguous range；
The average being calculated in upper step is subtracted into compensation constant Q, the threshold value of the pixel is obtained, in a kind of embodiment
In, Q is between [2,7]；
The pixel value is made comparisons with the threshold value that upper step is tried to achieve, if the pixel value is more than the threshold value, in the picture should
Pixel value is arranged to 255, and the pixel value otherwise is arranged into 0 in the picture.Binary conversion treatment is carried out to Fig. 3 and Fig. 8 image
Afterwards, it can obtain corresponding binary image as shown in figures 4 and 9.
S205：All profiles of binary image are classified and remove non-textual class profile, to determine text profile,
Each profile in binary image is proceeded as follows：
Determine the number nonz of the non-zero pixels value in profile；
Determine the depth-width ratio hw and the ratio of width to height wh of profile；
Determine in the range of the lateral contour of profile in horizontal contiguous range (it should be understood that laterally could alternatively be longitudinal direction or appoint
Anticipate suitable direction) existing for the profile number SH of similar width profile number SW and similar height；
If nonz is less than first threshold, and/or hw is more than Second Threshold and SH is less than the 3rd threshold value, and/or wh is more than
4th threshold value and SW is less than the 5th threshold value, then it is assumed that above-mentioned profile is all non-textual profile, by profile model in binary image
The pixel value for enclosing interior all positions is set to 0.In one embodiment, first threshold is between [2,5], Second Threshold [8,
12] between, the 3rd threshold value is between [2,5], and the 4th threshold value is between [8,12], and the 5th threshold value is between [2,5].Width
For W and height be H profile similar width outline definition be the profile width scope between [0.7W, 1.3W], it is similar
The profile elevations h of height profile is between [0.7H, 1.3H].
S206：Based on identified text profile, line of text is merged, to determine line of text region；
In this step, specifically, by the profile in the binary image after step S205 processing in the horizontal direction
(it should be understood that can also be in vertical direction or on any appropriate direction) carries out expansive working and etching operation, obtains text
Row region.Wherein, expansive working is to enter row bound addition to operation object, and it is then some pictures for deleting object bounds to corrode
Element, wherein, the definition on border is provided by corresponding operation operator, as Expanded Operators size be 5 × 1, then centered on this pixel,
Pixel in 5 × 1 contiguous range is disposed as object pixel, step S204, after the completion of S205, it is 5 × 1 to recycle size
Expanded Operators carry out 20 expansive workings, and using size be 5 × 1 erosion operator progress 15 etching operations, with institute
The boundary rectangle for having profile is mask, can obtain the line of text region of image, it should be appreciated that in concrete operations, art technology
Personnel can do appropriate adjustment to Expanded Operators, expansive working number and corrosion number.Fig. 5 and Figure 10 is that image to be assessed is complete
The line of text region obtained after into step S206.
Step S207：Quality evaluation is carried out to the line of text region detected in S106.
Before quality evaluation step, using one text-oriented row region of method training in advance based on deep learning
Image quality measure model.Wherein, the line of text data of training oneself can synthesize, and directly can also be intercepted from text image
Mark.Commonly used first from conventional Chinese individual character, English word and Chinese and English from the line of text of synthesis random in punctuation mark
Choose candidate characters composition character string, then blend character string and different background image, then add it is different degrees of obscure,
It is re-compressed into different degrees of quality and preservation.According to the mass parameter of composograph, the quality of correspondence image is marked, size is
Between 0-100, Fig. 6 is a schematic diagram of composograph.
4 line of text regions in the 10 line of text regions and Figure 10 in Fig. 5 are input to respectively train towards
The image quality measure model in line of text region, obtain corresponding line of text regional quality assessment result.Fig. 7 shows Fig. 5's
All line of text region line_1, line_2, line_3, line_4, line_5, line_6, line_7, line_8, line_9
With line_10 quality assessment result, Figure 11 shows Figure 10 all line of text region line_1a, line_2a, line_3a
With line_4a quality assessment result.
Step S108：Based on the quality assessment result in the line of text region obtained in S107, pass through mass-weighted average
Mode carries out quality evaluation to image to be assessed；
Wherein, weighted value determines according to the importance of target area, in this embodiment it is assumed that target area is important
Property it is identical, then average weighted mode is reduced to the mode of arithmetic average, and the quality assessment result of Fig. 3 text image is figure
All line of text region line_1, line_2, line_3, line_4, line_5, line_6, line_7, line_8 shown in 6,
The average value of line_9 and line_10 quality assessment result, size 16, the quality assessment result of Fig. 8 text image are
The quality assessment result of all line of text region line_1a, line_2a, line_3a and line_4a shown in Figure 11 are averaged
Value, size 97.
By foregoing description as can be seen that using above-described embodiment method for evaluating quality, pay close attention to target area interested
Domain, estimating velocity is fast, and Evaluation accuracy is higher, effectively assesses the quality of text image, the convenient text image different to quality
Do subsequent treatment.
In another embodiment, the flow of quality evaluation can equally include step S101, S102 and S103.It is specific and
Speech, can use the algorithm based on deep learning in target detection step.Can be with the step of target area quality evaluation
Using based on the statistics to grey scale pixel value come to target area carry out quality evaluation, for example, using Laplce's variance algorithm.
The quality assessment result based on target area can be used in the step of carrying out quality evaluation to image to be assessed, passes through quality
Average weighted mode carries out quality evaluation to image to be assessed.
The schematic diagram of image quality measure device according to embodiments of the present invention Figure 12.
Present invention also offers a kind of quality assessment device 1200 as shown in figure 12, the device includes object detection unit
1201st, target area quality estimation unit 1202 and image quality measure unit 1203 to be assessed.Specifically, object detection unit
1201 are configured as detecting image to be assessed, with true target area.Target area quality estimation unit 1202 is configured
To carry out quality evaluation to the target area determined in object detection unit 1201, to determine the quality evaluation knot of target area
Fruit.Image quality measure unit 1203 to be assessed is configured as based on identified mesh in target area quality estimation unit 1202
The quality assessment result in region is marked, quality evaluation is carried out to image to be assessed.For example, image quality measure unit 1203 to be assessed
Quality evaluation can be carried out to image to be assessed by way of mass-weighted average.
In addition, object detection unit 1201 includes profile sort module 1202a and text merging module 1202b.Profile point
Class unit 1202a is configured as based on determined by following at least one operation parameter to enter to the profile of the binary image
Row classification, to determine text profile：Determine the non-zero pixels number nonz in the profile of binary image；Determine binary image
Profile depth-width ratio hw and the ratio of width to height wh；And the designated parties in the range of the assigned direction of the profile of determination binary image
The number SH of the number SW of existing similar width profile and similar height profile into contiguous range.Text combining unit
1202b is additionally configured to by setting Expanded Operators and erosion operator, and identified text profile is carried out in the direction indicated
Expansive working and etching operation.Text combining unit 1202b is configured as based on identified text in profile taxon 1202a
This profile, text is merged in the direction indicated, to determine target area.
In one embodiment, target area quality estimation unit 1202 is additionally configured to the target area based on determined by
Domain and image quality measure model, determine the quality assessment result of target area.
In another embodiment, target area quality estimation unit 1202 is additionally configured to be based on to grey scale pixel value
Statistics come to target area carry out quality evaluation.
The flow of method for evaluating quality in Fig. 1,2 also represents machine readable instructions, the machine readable instructions include by
Manage the program that device performs.The program can be by hypostazation in the software for being stored in tangible computer computer-readable recording medium, the tangible meter
Calculation machine computer-readable recording medium such as CD-ROM, floppy disk, hard disk, digital versatile disc (DVD), the memory of Blu-ray Disc or other forms.
Substitute, some steps or all steps in the exemplary method in Fig. 1 can utilize application specific integrated circuit (ASIC), may be programmed and patrol
Any combination for collecting device (PLD), field programmable logic device (EPLD), discrete logic, hardware, firmware etc. is implemented.Separately
Outside, although the flow chart shown in Fig. 1 describes the method for evaluating quality, the step in the method for evaluating quality can be repaiied
Change, delete or merge.
As described above, realizing Fig. 1 instantiation procedure using coded command (such as computer-readable instruction), the programming refers to
Order is stored on tangible computer computer-readable recording medium, such as hard disk, flash memory, read-only storage (ROM), CD (CD), digital universal light
Disk (DVD), Cache, random access storage device (RAM) and/or any other storage medium, believe on the storage medium
Breath can store random time (for example, for a long time, for good and all, of short duration situation is interim to buffer, and/or the caching of information).Such as
As used herein, the term tangible computer computer-readable recording medium is expressly defined to include any type of computer-readable storage
Signal.Additionally or alternatively, Fig. 1 instantiation procedure, the coding are realized using coded command (such as computer-readable instruction)
Instruction is stored in non-transitory computer-readable medium, such as hard disk, flash memory, read-only storage, CD, digital versatile disc, height
Fast buffer, random access storage device and/or any other storage medium, random time can be stored in the storage-medium information
(for example, for a long time, for good and all, of short duration situation, interim buffering, and/or the caching of information).
Although describing the present invention with reference to specific example, wherein these specific examples are merely intended to be exemplary
, rather than limit the invention, but it will be apparent to those skilled in the art that do not departing from this
On the basis of the spirit and scope of invention, the disclosed embodiments can be changed, increased or deleted.
Claims (24)
- A kind of 1. image quality measure method, it is characterised in that including：Target detection is carried out to image to be assessed, to determine at least one target area；Quality evaluation is carried out respectively at least one target area, it is each at least one target area to determine The quality assessment result of target area；Based on the quality assessment result of each target area in identified at least one target area, to described to be evaluated Estimate image and carry out quality evaluation.
- 2. image quality measure method according to claim 1, it is characterised in that target is carried out to the image to be assessed The step of detection, includes：The profile of binary image to being obtained from the image to be assessed is classified, to determine text profile；Based on identified text profile, text is merged in the direction indicated, to determine at least one target area Domain.
- 3. image quality measure method according to claim 2, it is characterised in that based on following at least one operation institute really Fixed parameter is classified to the profile of the binary image：Determine the non-zero pixels number in the profile of the binary image；Determine the depth-width ratio and the ratio of width to height of the profile of the binary image；Determine existing similar width in the assigned direction contiguous range in the range of the assigned direction of the profile of the binary image Spend the number of profile and the number of similar height profile.
- 4. image quality measure method according to claim 2, it is characterised in that based on identified text profile, The step of being merged on assigned direction to text includes：By setting Expanded Operators and erosion operator, expansive working and corruption are carried out to identified text profile in the direction indicated Erosion operation.
- 5. image quality measure method according to claim 1, it is characterised in that at least one target area point It carry out not include the step of quality evaluation：Based on identified at least one target area and image quality measure model, at least one target area is determined The quality assessment result of each target area in domain.
- 6. image quality measure method according to claim 1, it is characterised in that at least one target area point It carry out not include the step of quality evaluation：Commented based on the statistics to grey scale pixel value to carry out quality to each target area at least one target area Estimate.
- 7. image quality measure method according to claim 1, it is characterised in that based on identified described at least one The quality assessment result of each target area in target area, the step of quality evaluation is carried out to the image to be assessed bag Include：By way of mass-weighted average, based on each target area in identified at least one target area Quality assessment result, quality evaluation is carried out to the image to be assessed.
- 8. image quality measure method according to claim 1, it is characterised in that at least one target area point Quality evaluation is not carried out, the step of to determine the quality assessment result of each target area at least one target area Including：Using reference-free quality evaluation index, quality evaluation is carried out respectively at least one target area, with described in determination The quality assessment result of each target area at least one target area,Wherein, the reference-free quality evaluation index includes at least one in edge strength, noise rate or uniform brightness distribution.
- A kind of 9. image quality measure device, it is characterised in that including：Object detection unit, it is configured as carrying out target detection to image to be assessed, to determine at least one target area；Target area quality estimation unit, it is configured as carrying out quality evaluation respectively at least one target area, with true The quality assessment result of each target area in fixed at least one target area；Image quality measure unit to be assessed, it is configured as based on each mesh in identified at least one target area The quality assessment result in region is marked, quality evaluation is carried out to the image to be assessed.
- 10. image quality measure device according to claim 9, it is characterised in that the object detection unit includes：Profile taxon, the profile for being configured as the binary image to being obtained from the image to be assessed are classified, with Determine text profile；Text combining unit, the text profile based on determined by is configured as, text is merged in the direction indicated, with true Fixed at least one target area.
- 11. image quality measure device according to claim 10, it is characterised in that the profile taxon also by with It is set to：The parameter based on determined by following at least one operation is classified to the profile of the binary image：Determine the non-zero pixels number in the profile of the binary image；Determine the depth-width ratio and the ratio of width to height of the profile of the binary image；Determine existing similar width in the assigned direction contiguous range in the range of the assigned direction of the profile of the binary image Spend the number of profile and the number of similar height profile.
- 12. image quality measure device according to claim 10, it is characterised in that the text combining unit also by with It is set to：By setting Expanded Operators and erosion operator, in the direction indicated to identified text profile carry out expansive working and Etching operation.
- 13. image quality measure device according to claim 9, it is characterised in that the target area quality evaluation list Member is additionally configured to：Based on identified at least one target area and image quality measure model, it is determined that it is described at least The quality assessment result of each target area in one target area.
- 14. image quality measure device according to claim 9, it is characterised in that the target area quality evaluation list Member is additionally configured to：Based on counting to enter to each target area at least one target area to grey scale pixel value Row quality evaluation.
- 15. image quality measure device according to claim 9, it is characterised in that the image quality measure to be assessed Unit is additionally configured to：By way of mass-weighted average, based on each target area in identified at least one target area Quality assessment result, quality evaluation is carried out to the image to be assessed.
- 16. image quality measure device according to claim 9, it is characterised in that the target area quality evaluation list Member is additionally configured to：Using reference-free quality evaluation index, quality evaluation is carried out respectively at least one target area, with The quality assessment result of each target area at least one target area is determined, wherein, the no reference mass is commented Valency index includes at least one in edge strength, noise rate or uniform brightness distribution.
- 17. a kind of computer-readable storage medium, be stored thereon with processor executable program, when can be held described in the computing device During line program, follow the steps below：.Target detection is carried out to image to be assessed, to determine at least one target area；Quality evaluation is carried out respectively at least one target area, it is each at least one target area to determine The quality assessment result of target area；Based on the quality assessment result of each target area in identified at least one target area, to described to be evaluated Estimate image and carry out quality evaluation.
- 18. computer-readable storage medium according to claim 17, it is characterised in that when can be held described in the computing device During line program, the step of image progress target detection to be assessed, is included：The profile of binary image to being obtained from the image to be assessed is classified, to determine text profile；Based on identified text profile, text is merged in the direction indicated, to determine at least one target area Domain.
- 19. computer-readable storage medium according to claim 18, it is characterised in that when can be held described in the computing device During line program, the parameter based on determined by following at least one operation is classified to the profile of the binary image：Determine the non-zero pixels number in the profile of the binary image；Determine the depth-width ratio and the ratio of width to height of the profile of the binary image；Determine existing similar width in the assigned direction contiguous range in the range of the assigned direction of the profile of the binary image Spend the number of profile and the number of similar height profile.
- 20. computer-readable storage medium according to claim 18, it is characterised in that when can be held described in the computing device During line program, based on identified text profile, the step of being merged in the direction indicated to text, includes：By setting Expanded Operators and erosion operator, expansive working and corruption are carried out to identified text profile in the direction indicated Erosion operation.
- 21. computer-readable storage medium according to claim 17, it is characterised in that when can be held described in the computing device During line program, the step of carrying out quality evaluation respectively at least one target area, includes：Based on identified at least one target area and image quality measure model, at least one target area is determined The quality assessment result of each target area in domain.
- 22. computer-readable storage medium according to claim 17, it is characterised in that when can be held described in the computing device During line program, the step of carrying out quality evaluation respectively at least one target area, includes：Commented based on the statistics to grey scale pixel value to carry out quality to each target area at least one target area Estimate.
- 23. computer-readable storage medium according to claim 17, it is characterised in that when can be held described in the computing device During line program, based on the quality assessment result of each target area in identified at least one target area, to institute Stating the step of image to be assessed carries out quality evaluation includes：By way of mass-weighted average, based on each target area in identified at least one target area Quality assessment result, quality evaluation is carried out to the image to be assessed.
- 24. computer-readable storage medium according to claim 17, it is characterised in that when can be held described in the computing device During line program, quality evaluation is carried out respectively at least one target area, to determine at least one target area Each target area quality assessment result the step of include：Using reference-free quality evaluation index, quality evaluation is carried out respectively at least one target area, with described in determination The quality assessment result of each target area at least one target area, wherein, the reference-free quality evaluation index bag Include at least one in edge strength, noise rate or uniform brightness distribution.
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710854415.9A CN107481238A (en) | 2017-09-20 | 2017-09-20 | Image quality measure method and device |
| SG11201907815V SG11201907815VA (en) | 2017-09-20 | 2018-09-19 | Method for assessing image quality and device thereof |
| JP2020504760A JP2020513133A (en) | 2017-09-20 | 2018-09-19 | Image quality evaluation method and apparatus |
| PCT/CN2018/106451 WO2019057067A1 (en) | 2017-09-20 | 2018-09-19 | Image quality evaluation method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710854415.9A CN107481238A (en) | 2017-09-20 | 2017-09-20 | Image quality measure method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107481238A true CN107481238A (en) | 2017-12-15 |
Family
ID=60586643
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710854415.9A Pending CN107481238A (en) | 2017-09-20 | 2017-09-20 | Image quality measure method and device |
Country Status (4)
| Country | Link |
|---|---|
| JP (1) | JP2020513133A (en) |
| CN (1) | CN107481238A (en) |
| SG (1) | SG11201907815VA (en) |
| WO (1) | WO2019057067A1 (en) |
Cited By (32)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108122231A (en) * | 2018-01-10 | 2018-06-05 | 山东华软金盾软件股份有限公司 | Image quality evaluating method based on ROI Laplacian algorithms under monitor video |
| CN108460766A (en) * | 2018-04-12 | 2018-08-28 | 四川和生视界医药技术开发有限公司 | A kind of retinal images intelligibility evaluation method and apparatus for evaluating |
| CN108596084A (en) * | 2018-04-23 | 2018-09-28 | 宁波Gqy视讯股份有限公司 | A kind of charging pile automatic identifying method and device |
| CN108805172A (en) * | 2018-05-08 | 2018-11-13 | 重庆瑞景信息科技有限公司 | A kind of blind evaluation method of image efficiency of object-oriented |
| CN108875731A (en) * | 2017-12-28 | 2018-11-23 | 北京旷视科技有限公司 | Target identification method, device, system and storage medium |
| CN109104568A (en) * | 2018-07-24 | 2018-12-28 | 苏州佳世达光电有限公司 | The intelligent cleaning driving method and drive system of monitoring camera |
| WO2019057067A1 (en) * | 2017-09-20 | 2019-03-28 | 众安信息技术服务有限公司 | Image quality evaluation method and apparatus |
| CN109948625A (en) * | 2019-03-07 | 2019-06-28 | 上海汽车集团股份有限公司 | Definition of text images appraisal procedure and system, computer readable storage medium |
| CN110245577A (en) * | 2019-05-23 | 2019-09-17 | 复钧智能科技(苏州)有限公司 | Target vehicle recognition methods, device and Vehicular real time monitoring system |
| WO2019174130A1 (en) * | 2018-03-14 | 2019-09-19 | 平安科技（深圳）有限公司 | Bill recognition method, server, and computer readable storage medium |
| CN110287826A (en) * | 2019-06-11 | 2019-09-27 | 北京工业大学 | A kind of video object detection method based on attention mechanism |
| CN110414519A (en) * | 2019-06-27 | 2019-11-05 | 众安信息技术服务有限公司 | A kind of recognition methods of picture character and its identification device |
| CN110827261A (en) * | 2019-11-05 | 2020-02-21 | 泰康保险集团股份有限公司 | Image quality detection method and device, storage medium and electronic equipment |
| CN110874547A (en) * | 2018-08-30 | 2020-03-10 | 富士通株式会社 | Method and device for identifying object from video |
| CN111192241A (en) * | 2019-12-23 | 2020-05-22 | 深圳市优必选科技股份有限公司 | Quality evaluation method and device of face image and computer storage medium |
| CN111368837A (en) * | 2018-12-25 | 2020-07-03 | 中移（杭州）信息技术有限公司 | Image quality evaluation method and device, electronic equipment and storage medium |
| CN111417981A (en) * | 2018-03-12 | 2020-07-14 | 华为技术有限公司 | Image definition detection method and device |
| CN111595267A (en) * | 2020-05-18 | 2020-08-28 | 浙江大华技术股份有限公司 | Method, device, storage medium and electronic device for determining phase value of object |
| CN112102309A (en) * | 2020-09-27 | 2020-12-18 | 中国建设银行股份有限公司 | Method, device and equipment for determining image quality evaluation result |
| CN112396050A (en) * | 2020-12-02 | 2021-02-23 | 上海优扬新媒信息技术有限公司 | Image processing method, device and storage medium |
| CN112396574A (en) * | 2019-08-02 | 2021-02-23 | 浙江宇视科技有限公司 | License plate image quality processing method and device, storage medium and electronic equipment |
| WO2021082819A1 (en) * | 2019-10-31 | 2021-05-06 | 北京金山云网络技术有限公司 | Image generation method and apparatus, and electronic device |
| CN112801132A (en) * | 2020-12-28 | 2021-05-14 | 泰康保险集团股份有限公司 | Image processing method and device |
| CN112991313A (en) * | 2021-03-29 | 2021-06-18 | 清华大学 | Image quality evaluation method and device, electronic device and storage medium |
| CN113506260A (en) * | 2021-07-05 | 2021-10-15 | 北京房江湖科技有限公司 | Face image quality evaluation method and device, electronic equipment and storage medium |
| CN113537192A (en) * | 2021-06-30 | 2021-10-22 | 北京百度网讯科技有限公司 | Image detection method, image detection device, electronic equipment and storage medium |
| CN113627419A (en) * | 2020-05-08 | 2021-11-09 | 百度在线网络技术（北京）有限公司 | Interest region evaluation method, device, equipment and medium |
| CN113781428A (en) * | 2021-09-09 | 2021-12-10 | 广东电网有限责任公司 | Image processing method and device, electronic equipment and storage medium |
| CN114067006A (en) * | 2022-01-17 | 2022-02-18 | 湖南工商大学 | Screen content image quality evaluation method based on discrete cosine transform |
| CN114219803A (en) * | 2022-02-21 | 2022-03-22 | 浙江大学 | Detection method and system for three-stage image quality evaluation |
| CN114387254A (en) * | 2022-01-12 | 2022-04-22 | 中国平安人寿保险股份有限公司 | Document quality analysis method and device, computer equipment and storage medium |
| WO2024119322A1 (en) * | 2022-12-05 | 2024-06-13 | 深圳华大生命科学研究院 | Method and apparatus for evaluating quality of grayscale image, and electronic device and storage medium |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110236544B (en) * | 2019-05-29 | 2023-05-02 | 中国科学院重庆绿色智能技术研究院 | Stroke perfusion imaging lesion area detection system and method based on correlation coefficient |
| CN110232381B (en) * | 2019-06-19 | 2023-06-20 | 梧州学院 | License plate segmentation method, license plate segmentation device, computer equipment and computer readable storage medium |
| CN111046886B (en) * | 2019-12-12 | 2023-05-12 | 吉林大学 | Automatic identification method, device and equipment for number plate and computer readable storage medium |
| CN111192258A (en) * | 2020-01-02 | 2020-05-22 | 广州大学 | Image quality evaluation method and device |
| CN111275681B (en) * | 2020-01-19 | 2023-09-01 | 浙江大华技术股份有限公司 | Picture quality determining method and device, storage medium and electronic device |
| CN111696083B (en) * | 2020-05-20 | 2024-05-14 | 平安科技（深圳）有限公司 | Image processing method and device, electronic equipment and storage medium |
| CN112365451B (en) * | 2020-10-23 | 2024-06-21 | 微民保险代理有限公司 | Method, device, equipment and computer readable medium for determining image quality grade |
| CN112287898B (en) * | 2020-11-26 | 2024-07-05 | 深源恒际科技有限公司 | Method and system for evaluating text detection quality of image |
| CN112767318B (en) * | 2020-12-31 | 2023-07-25 | 科大讯飞股份有限公司 | Image processing effect evaluation method, device, storage medium and equipment |
| CN113450323B (en) * | 2021-06-22 | 2022-12-06 | 深圳盈天下视觉科技有限公司 | Quality detection method and device, electronic equipment and computer readable storage medium |
| CN113724196A (en) * | 2021-07-16 | 2021-11-30 | 北京工业大学 | Image quality evaluation method, device, equipment and storage medium |
| CN115239961B (en) * | 2022-09-21 | 2022-12-20 | 江苏跃格智能装备有限公司 | Method for monitoring working state of laser cutting machine |
| CN116246273B (en) * | 2023-03-07 | 2024-03-22 | 广州市易鸿智能装备有限公司 | Image annotation consistency evaluation method and device, electronic equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101533474A (en) * | 2008-03-12 | 2009-09-16 | 三星电子株式会社 | Character and image recognition system based on video image and method thereof |
| CN102054271A (en) * | 2009-11-02 | 2011-05-11 | 富士通株式会社 | Text line detection method and device |
| US20160283787A1 (en) * | 2008-01-18 | 2016-09-29 | Mitek Systems, Inc. | Systems and methods for mobile image capture and content processing of driver's licenses |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3400151B2 (en) * | 1994-12-08 | 2003-04-28 | 株式会社東芝 | Character string region extraction apparatus and method |
| JP3710164B2 (en) * | 1995-05-02 | 2005-10-26 | キヤノン株式会社 | Image processing apparatus and method |
| JP2003208568A (en) * | 2002-01-10 | 2003-07-25 | Ricoh Co Ltd | Image processor, image processing method and program used in the method |
| JP2007156741A (en) * | 2005-12-02 | 2007-06-21 | Koito Ind Ltd | Character extraction method, character extraction device, and program |
| JP4821869B2 (en) * | 2009-03-18 | 2011-11-24 | 富士ゼロックス株式会社 | Character recognition device, image reading device, and program |
| CN103049893B (en) * | 2011-10-14 | 2015-12-16 | 深圳信息职业技术学院 | A kind of method of image fusion quality assessment and device |
| CN106686377B (en) * | 2016-12-30 | 2018-09-04 | 佳都新太科技股份有限公司 | A kind of video emphasis area determination method based on deep-neural-network |
| CN107123122B (en) * | 2017-04-28 | 2020-06-12 | 深圳大学 | No-reference image quality evaluation method and device |
| CN107481238A (en) * | 2017-09-20 | 2017-12-15 | 众安信息技术服务有限公司 | Image quality measure method and device |
-
2017
- 2017-09-20 CN CN201710854415.9A patent/CN107481238A/en active Pending
-
2018
- 2018-09-19 WO PCT/CN2018/106451 patent/WO2019057067A1/en active Application Filing
- 2018-09-19 JP JP2020504760A patent/JP2020513133A/en active Pending
- 2018-09-19 SG SG11201907815V patent/SG11201907815VA/en unknown
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160283787A1 (en) * | 2008-01-18 | 2016-09-29 | Mitek Systems, Inc. | Systems and methods for mobile image capture and content processing of driver's licenses |
| CN101533474A (en) * | 2008-03-12 | 2009-09-16 | 三星电子株式会社 | Character and image recognition system based on video image and method thereof |
| CN102054271A (en) * | 2009-11-02 | 2011-05-11 | 富士通株式会社 | Text line detection method and device |
Non-Patent Citations (2)
| Title |
|---|
| 周景超: "视频文本检测算法研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
| 王志明: "无参考图像质量评价综述", 《自动化学报》 * |
Cited By (48)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019057067A1 (en) * | 2017-09-20 | 2019-03-28 | 众安信息技术服务有限公司 | Image quality evaluation method and apparatus |
| CN108875731A (en) * | 2017-12-28 | 2018-11-23 | 北京旷视科技有限公司 | Target identification method, device, system and storage medium |
| CN108122231B (en) * | 2018-01-10 | 2021-09-24 | 山东华软金盾软件股份有限公司 | Image quality evaluation method based on ROI Laplacian algorithm under monitoring video |
| CN108122231A (en) * | 2018-01-10 | 2018-06-05 | 山东华软金盾软件股份有限公司 | Image quality evaluating method based on ROI Laplacian algorithms under monitor video |
| CN111417981A (en) * | 2018-03-12 | 2020-07-14 | 华为技术有限公司 | Image definition detection method and device |
| WO2019174130A1 (en) * | 2018-03-14 | 2019-09-19 | 平安科技（深圳）有限公司 | Bill recognition method, server, and computer readable storage medium |
| CN108460766B (en) * | 2018-04-12 | 2022-02-25 | 四川和生视界医药技术开发有限公司 | Retina image definition evaluation method and evaluation device |
| CN108460766A (en) * | 2018-04-12 | 2018-08-28 | 四川和生视界医药技术开发有限公司 | A kind of retinal images intelligibility evaluation method and apparatus for evaluating |
| CN108596084A (en) * | 2018-04-23 | 2018-09-28 | 宁波Gqy视讯股份有限公司 | A kind of charging pile automatic identifying method and device |
| CN108805172A (en) * | 2018-05-08 | 2018-11-13 | 重庆瑞景信息科技有限公司 | A kind of blind evaluation method of image efficiency of object-oriented |
| CN109104568A (en) * | 2018-07-24 | 2018-12-28 | 苏州佳世达光电有限公司 | The intelligent cleaning driving method and drive system of monitoring camera |
| CN110874547A (en) * | 2018-08-30 | 2020-03-10 | 富士通株式会社 | Method and device for identifying object from video |
| CN110874547B (en) * | 2018-08-30 | 2023-09-12 | 富士通株式会社 | Method and apparatus for identifying objects from video |
| CN111368837B (en) * | 2018-12-25 | 2023-12-05 | 中移（杭州）信息技术有限公司 | Image quality evaluation method and device, electronic equipment and storage medium |
| CN111368837A (en) * | 2018-12-25 | 2020-07-03 | 中移（杭州）信息技术有限公司 | Image quality evaluation method and device, electronic equipment and storage medium |
| CN109948625A (en) * | 2019-03-07 | 2019-06-28 | 上海汽车集团股份有限公司 | Definition of text images appraisal procedure and system, computer readable storage medium |
| CN110245577A (en) * | 2019-05-23 | 2019-09-17 | 复钧智能科技(苏州)有限公司 | Target vehicle recognition methods, device and Vehicular real time monitoring system |
| CN110287826A (en) * | 2019-06-11 | 2019-09-27 | 北京工业大学 | A kind of video object detection method based on attention mechanism |
| CN110287826B (en) * | 2019-06-11 | 2021-09-17 | 北京工业大学 | Video target detection method based on attention mechanism |
| CN110414519B (en) * | 2019-06-27 | 2023-11-14 | 众安信息技术服务有限公司 | Picture character recognition method and device and storage medium |
| CN110414519A (en) * | 2019-06-27 | 2019-11-05 | 众安信息技术服务有限公司 | A kind of recognition methods of picture character and its identification device |
| CN112396574A (en) * | 2019-08-02 | 2021-02-23 | 浙江宇视科技有限公司 | License plate image quality processing method and device, storage medium and electronic equipment |
| CN112396574B (en) * | 2019-08-02 | 2024-02-02 | 浙江宇视科技有限公司 | License plate image quality processing method and device, storage medium and electronic equipment |
| US11836898B2 (en) | 2019-10-31 | 2023-12-05 | Beijing Kingsoft Cloud Network Technology Co., Ltd. | Method and apparatus for generating image, and electronic device |
| WO2021082819A1 (en) * | 2019-10-31 | 2021-05-06 | 北京金山云网络技术有限公司 | Image generation method and apparatus, and electronic device |
| CN110827261A (en) * | 2019-11-05 | 2020-02-21 | 泰康保险集团股份有限公司 | Image quality detection method and device, storage medium and electronic equipment |
| CN111192241B (en) * | 2019-12-23 | 2024-02-13 | 深圳市优必选科技股份有限公司 | Quality evaluation method and device for face image and computer storage medium |
| CN111192241A (en) * | 2019-12-23 | 2020-05-22 | 深圳市优必选科技股份有限公司 | Quality evaluation method and device of face image and computer storage medium |
| CN113627419A (en) * | 2020-05-08 | 2021-11-09 | 百度在线网络技术（北京）有限公司 | Interest region evaluation method, device, equipment and medium |
| CN111595267A (en) * | 2020-05-18 | 2020-08-28 | 浙江大华技术股份有限公司 | Method, device, storage medium and electronic device for determining phase value of object |
| CN112102309A (en) * | 2020-09-27 | 2020-12-18 | 中国建设银行股份有限公司 | Method, device and equipment for determining image quality evaluation result |
| CN112396050A (en) * | 2020-12-02 | 2021-02-23 | 上海优扬新媒信息技术有限公司 | Image processing method, device and storage medium |
| CN112396050B (en) * | 2020-12-02 | 2023-09-15 | 度小满科技(北京)有限公司 | Image processing method, device and storage medium |
| CN112801132A (en) * | 2020-12-28 | 2021-05-14 | 泰康保险集团股份有限公司 | Image processing method and device |
| CN112801132B (en) * | 2020-12-28 | 2024-01-02 | 泰康同济(武汉)医院 | Image processing method and device |
| CN112991313B (en) * | 2021-03-29 | 2021-09-14 | 清华大学 | Image quality evaluation method and device, electronic device and storage medium |
| CN112991313A (en) * | 2021-03-29 | 2021-06-18 | 清华大学 | Image quality evaluation method and device, electronic device and storage medium |
| CN113537192A (en) * | 2021-06-30 | 2021-10-22 | 北京百度网讯科技有限公司 | Image detection method, image detection device, electronic equipment and storage medium |
| CN113537192B (en) * | 2021-06-30 | 2024-03-26 | 北京百度网讯科技有限公司 | Image detection method, device, electronic equipment and storage medium |
| CN113506260B (en) * | 2021-07-05 | 2023-08-29 | 贝壳找房(北京)科技有限公司 | Face image quality assessment method and device, electronic equipment and storage medium |
| CN113506260A (en) * | 2021-07-05 | 2021-10-15 | 北京房江湖科技有限公司 | Face image quality evaluation method and device, electronic equipment and storage medium |
| CN113781428A (en) * | 2021-09-09 | 2021-12-10 | 广东电网有限责任公司 | Image processing method and device, electronic equipment and storage medium |
| CN114387254A (en) * | 2022-01-12 | 2022-04-22 | 中国平安人寿保险股份有限公司 | Document quality analysis method and device, computer equipment and storage medium |
| CN114067006B (en) * | 2022-01-17 | 2022-04-08 | 湖南工商大学 | Screen content image quality evaluation method based on discrete cosine transform |
| CN114067006A (en) * | 2022-01-17 | 2022-02-18 | 湖南工商大学 | Screen content image quality evaluation method based on discrete cosine transform |
| CN114219803B (en) * | 2022-02-21 | 2022-07-15 | 浙江大学 | Detection method and system for three-stage image quality evaluation |
| CN114219803A (en) * | 2022-02-21 | 2022-03-22 | 浙江大学 | Detection method and system for three-stage image quality evaluation |
| WO2024119322A1 (en) * | 2022-12-05 | 2024-06-13 | 深圳华大生命科学研究院 | Method and apparatus for evaluating quality of grayscale image, and electronic device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| SG11201907815VA (en) | 2019-11-28 |
| JP2020513133A (en) | 2020-04-30 |
| WO2019057067A1 (en) | 2019-03-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107481238A (en) | Image quality measure method and device | |
| CN109902677B (en) | Vehicle detection method based on deep learning | |
| CN110738101B (en) | Behavior recognition method, behavior recognition device and computer-readable storage medium | |
| US20190188528A1 (en) | Text detection method and apparatus, and storage medium | |
| KR102114357B1 (en) | Method and device for constructing a table including information on a pooling type and testing method and testing device using the same | |
| CN110874841A (en) | Object detection method and device with reference to edge image | |
| CN111259878A (en) | Method and equipment for detecting text | |
| JP6698191B1 (en) | Road marking failure detection device, road marking failure detection method, and road marking failure detection program | |
| KR101652261B1 (en) | Method for detecting object using camera | |
| CN110599453A (en) | Panel defect detection method and device based on image fusion and equipment terminal | |
| CN117351011B (en) | Screen defect detection method, apparatus, and readable storage medium | |
| CN113537037A (en) | Pavement disease identification method, system, electronic device and storage medium | |
| CN111598884A (en) | Image data processing method, apparatus and computer storage medium | |
| CN114926407A (en) | Steel surface defect detection system based on deep learning | |
| CN106780727A (en) | A kind of headstock detection model method for reconstructing and device | |
| KR20200039043A (en) | Object recognition device and operating method for the same | |
| CN110070549A (en) | A kind of soft dividing method of marine oil overflow SAR image based on optimal scale neighborhood information | |
| CN111652140A (en) | Method, device, equipment and medium for accurately segmenting questions based on deep learning | |
| CN112733823A (en) | Method and device for extracting key frame for gesture recognition and readable storage medium | |
| CN115953744A (en) | Vehicle identification tracking method based on deep learning | |
| CN114419006A (en) | Method and system for removing watermark of gray level video characters changing along with background | |
| CN117037049B (en) | Image content detection method and system based on YOLOv5 deep learning | |
| JP2010002960A (en) | Image processor, image processing method, and image processing program | |
| JP4516994B2 (en) | Method and system for determining the background color of a digital image | |
| CN110287752A (en) | A kind of dot matrix code detection method and device |
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 | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 1248383 Country of ref document: HK |
|
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20171215 |
|
| WD01 | Invention patent application deemed withdrawn after publication |