CN117911930B - Data security early warning method and device based on intelligent video monitoring - Google Patents
Data security early warning method and device based on intelligent video monitoring Download PDFInfo
- Publication number
- CN117911930B CN117911930B CN202410294857.2A CN202410294857A CN117911930B CN 117911930 B CN117911930 B CN 117911930B CN 202410294857 A CN202410294857 A CN 202410294857A CN 117911930 B CN117911930 B CN 117911930B
- Authority
- CN
- China
- Prior art keywords
- video
- abnormal
- sequence
- classification
- frame
- 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.)
- Active
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 130
- 238000000034 method Methods 0.000 title claims abstract description 37
- 230000002159 abnormal effect Effects 0.000 claims abstract description 303
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 46
- 238000010586 diagram Methods 0.000 claims abstract description 29
- 238000001514 detection method Methods 0.000 claims description 22
- 238000012937 correction Methods 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000013507 mapping Methods 0.000 claims description 7
- 230000005856 abnormality Effects 0.000 claims description 2
- 238000013145 classification model Methods 0.000 claims 1
- 230000006870 function Effects 0.000 description 17
- 206010000117 Abnormal behaviour Diseases 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 4
- 238000003066 decision tree Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000002776 aggregation Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- 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
- G06V10/443—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 by matching or filtering
- G06V10/449—Biologically inspired filters, e.g. difference of Gaussians [DoG] or Gabor filters
- G06V10/451—Biologically inspired filters, e.g. difference of Gaussians [DoG] or Gabor filters with interaction between the filter responses, e.g. cortical complex cells
- G06V10/454—Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Databases & Information Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biodiversity & Conservation Biology (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Alarm Systems (AREA)
Abstract
The invention relates to the technical field of information security, and discloses a data security early warning method and device based on intelligent video monitoring, wherein the method comprises the following steps: generating a video interframe sequence diagram of the monitoring video image through the time attribute and the associated attribute; extracting video single frame characteristics of a video inter-frame sequence chart, and detecting abnormal data in the video single frame characteristics to obtain a video abnormal sequence; classifying and identifying the video abnormal sequence to obtain a video classified abnormal sequence, and determining an abnormal value of the video classified abnormal sequence according to the classification coefficient; calculating a sequence anomaly threshold value of the video anomaly sequence according to the anomaly value and the dynamic weight by a dynamic weight algorithm; and calculating a safety early warning value of the monitoring video image according to the sequence abnormal critical value, and determining the data safety of the monitoring video image according to the safety early warning value. The invention can improve the accuracy of data security early warning.
Description
Technical Field
The invention relates to the technical field of information security, in particular to a data security early warning method and device based on intelligent video monitoring.
Background
As an effective means of security protection, the video monitoring system is increasingly receiving attention, and the intelligent video monitoring technology is a brand new multifunctional video monitoring technology commonly applied to various markets, universities and public places, but in order to ensure the security of intelligent video monitoring data, abnormal data in the monitoring data needs to be analyzed so as to perform data security early warning.
The existing data safety early warning technology is mainly used for realizing mode detection of abnormal data or abnormal behaviors of semi-supervised learning by constructing a model. In practical application, the video monitoring data is changed in real time, only the supervision and detection of the historical data are considered, the data security early warning mode can be excessively one-sided, and the accuracy in the data security early warning process can be low.
Disclosure of Invention
The invention provides a data security early warning method and device based on intelligent video monitoring, and mainly aims to solve the problem of low accuracy in data security early warning.
In order to achieve the above purpose, the data security early warning method based on intelligent video monitoring provided by the invention comprises the following steps:
S1, acquiring a preset monitoring video image, and generating a video interframe sequence diagram of the monitoring video image through a preset time attribute and a preset association attribute;
s2, extracting video single-frame characteristics of the video inter-frame sequence diagram, and detecting abnormal data in the video single-frame characteristics by using a preset abnormal detection algorithm to obtain a video abnormal sequence;
S3, carrying out classification recognition on the video abnormal sequence through a preset classification recognition model to obtain a video classification abnormal sequence, and determining an abnormal value of the video classification abnormal sequence according to a preset classification coefficient;
s4, calculating a sequence anomaly threshold value of the video anomaly sequence according to the anomaly value and a preset dynamic weight through a preset dynamic weight algorithm, wherein the calculating the sequence anomaly threshold value of the video anomaly sequence according to the anomaly value and the preset dynamic weight through the preset dynamic weight algorithm comprises the following steps:
s41, dynamically monitoring the video abnormal sequence to obtain a dynamic video abnormal sequence;
S42, determining dynamic correction parameters of the dynamic video abnormal sequence;
s43, calculating a sequence anomaly threshold value of the video anomaly sequence according to the anomaly value, the dynamic weight and the dynamic correction parameter through the dynamic weight algorithm, wherein the dynamic weight algorithm is as follows:
；
Wherein, For the sequence anomaly threshold value,/>Classifying sequence numbers of abnormal sequences for video,/>For/>Dynamic weighting of individual video classification anomaly sequences,/>For/>Abnormal value of individual video classification abnormal sequence,/>For the dynamic correction parameters,/>As a logarithmic function;
And S5, calculating a safety early warning value of the monitoring video image according to the sequence abnormal critical value by using a preset safety early warning algorithm, and determining the data safety of the monitoring video image according to the safety early warning value.
Optionally, the generating the video inter-frame sequence chart of the monitoring video image through the preset time attribute and the preset association attribute includes:
generating a video frame node of the monitoring video image according to the time attribute;
Constructing an inter-frame relation between each video frame node according to the association attribute;
And associating the video frame nodes with the inter-frame relationship through a preset time sequence relationship to obtain a video inter-frame sequence diagram of the monitoring video image.
Optionally, the constructing an inter-frame relationship between each video frame node according to the association attribute includes:
mapping each video frame node into a distance space according to the association attribute;
Determining node edges from node distances between the video frame nodes in the distance space, wherein the node edges are defined as:
；
Wherein, Represents the/>Video frame node and/>Node edge value of each video frame node,/>Represents the/>Video frame node and/>Node distance between individual video frame nodes,/>Representing node effective distance,/>Represent the firstVideo frame node and/>Node distance between individual video frame nodes,/>Representing real number set,/>Representing an exponential function;
And determining the inter-frame relationship between each video frame node according to the node edges.
Optionally, the detecting the abnormal data in the video single frame feature by using a preset abnormal detection algorithm to obtain a video abnormal sequence includes:
Calculating the feature distance scale between each video frame and a preset normal frame one by one according to the video single frame features by using the anomaly detection algorithm;
When the characteristic distance scale is larger than a preset characteristic distance threshold, taking the video frame corresponding to the characteristic distance scale as abnormal data;
and collecting video frames corresponding to the abnormal data as the video abnormal sequence.
Optionally, before the classifying and identifying the video abnormal sequence through the preset classifying and identifying model, the method further includes:
Acquiring abnormal classification attributes of the monitoring video image;
Calculating the entropy value of each abnormal data in the preset video sample abnormal sequence according to the abnormal classification attribute by using the following entropy value calculation formula:
；
Wherein, Represents the/>Entropy of the abnormal data,/>Representing the amount of data in the abnormal sequence of video samples,/>Indicating that the anomaly data belongs to the/>Number of the abnormal classification attributes,/>Representing the classification quantity of the abnormal classification attribute,/>Representing a logarithmic function;
Selecting the abnormal data with the maximum entropy value as a branch node, marking the branch node as a classification attribute node, deleting the abnormal data corresponding to the branch node from the video sample abnormal sequence, and obtaining an updated video sample sequence;
When the classification attribute node exists, calculating an entropy value of each anomaly in the updated video sample sequence, and returning to the step of selecting the anomaly data with the maximum entropy value as a branch node until the classification attribute node does not exist;
And when the classification attribute node does not exist, generating the classification recognition model.
Optionally, the classifying and identifying the video abnormal sequence through a preset classifying and identifying model to obtain a video classifying and identifying abnormal sequence includes:
Inputting the abnormal data in the video abnormal sequence one by one into the classification recognition model to perform classification judgment, and obtaining classification attributes corresponding to the abnormal data;
and collecting the abnormal data with the same classification attribute as the video classification abnormal sequence.
Optionally, the determining the abnormal value of the video classification abnormal sequence according to the preset classification coefficient includes:
counting the total quantity of abnormal data in the video classification abnormal sequence;
Calculating the outlier of the video classification outlier sequence according to the classification coefficient and the total number by using the outlier calculation function as follows:
；
Wherein, For/>Abnormal value of individual video classification abnormal sequence,/>For/>Classification coefficient of abnormal sequence of individual video classification,/>For/>Data quantity of abnormal sequence of individual video classification,/>Is the total number.
Optionally, the calculating, by using a preset security early warning algorithm, the security early warning value of the surveillance video image according to the sequence anomaly threshold value includes:
determining the number of sequences with abnormal values higher than the sequence abnormal critical value corresponding to the video classification abnormal sequence by utilizing the safety early warning algorithm;
calculating abnormal frequency of the monitoring video image according to the number of the sequences;
And determining a safety precaution value of the monitoring video image according to the abnormal frequency.
Optionally, the determining the data security of the surveillance video image according to the security pre-warning value includes:
Determining the data security early warning level of the monitoring video image according to the security early warning value;
and triggering an alarm command for data security of the monitoring video image when the early warning level is higher than a preset security early warning level.
In order to solve the above problems, the present invention further provides a data security early warning device based on intelligent video monitoring, the device comprising:
The video inter-frame sequence diagram generation module is used for acquiring a preset monitoring video image and generating a video inter-frame sequence diagram of the monitoring video image through a preset time attribute and a preset association attribute;
The abnormal data detection module is used for extracting video single frame characteristics of the video inter-frame sequence diagram, and detecting abnormal data in the video single frame characteristics by using a preset abnormal detection algorithm to obtain a video abnormal sequence;
The abnormal value calculation module is used for carrying out classification recognition on the video abnormal sequence through a preset classification recognition model to obtain a video classification abnormal sequence, and determining an abnormal value of the video classification abnormal sequence according to a preset classification coefficient;
The sequence anomaly threshold value calculation module is used for calculating the sequence anomaly threshold value of the video anomaly sequence according to the anomaly value and the preset dynamic weight through a preset dynamic weight algorithm;
And the data safety early warning module is used for calculating the safety early warning value of the monitoring video image according to the sequence abnormal critical value by utilizing a preset safety early warning algorithm and determining the data safety of the monitoring video image according to the safety early warning value.
According to the embodiment of the invention, the video inter-frame sequence diagram of the monitoring video image is generated through the time attribute and the associated attribute, so that abnormal data is detected according to the video inter-frame sequence diagram to obtain a video abnormal sequence, each frame of image in the monitoring video is accurately detected, and the integrity of abnormal data detection is realized; classifying according to the video abnormal sequence, further determining an abnormal value of the video classified abnormal sequence, and calculating a sequence abnormal critical value of the video abnormal sequence according to the abnormal value and the dynamic weight, so that accurate analysis of the data security of the complete monitoring video is facilitated; and determining the safety precaution value of the monitoring video image according to the sequence abnormal critical value, and further determining the data safety of the monitoring video image according to the safety precaution value, thereby being beneficial to improving the accuracy of the data safety precaution. Therefore, the data security early warning method and device based on intelligent video monitoring can solve the problem of low accuracy of data security early warning.
Drawings
Fig. 1 is a schematic flow chart of a data security early warning method based on intelligent video monitoring according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a method for generating a video interframe sequence map according to an embodiment of the present invention;
FIG. 3 is a flow chart of detecting abnormal data according to an embodiment of the present invention;
FIG. 4 is a functional block diagram of a data security pre-warning device based on intelligent video monitoring according to an embodiment of the present invention;
The achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The embodiment of the application provides a data security early warning method based on intelligent video monitoring. The execution main body of the data security early warning method based on intelligent video monitoring comprises at least one of electronic equipment, such as a server side, a terminal and the like, which can be configured to execute the method provided by the embodiment of the application. In other words, the data security early warning method based on intelligent video monitoring can be executed by software or hardware installed in a terminal device or a server device, and the software can be a blockchain platform. The service end includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like. The server may be an independent server, or may be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.
Referring to fig. 1, a flow chart of a data security early warning method based on intelligent video monitoring according to an embodiment of the invention is shown. In this embodiment, the data security early warning method based on intelligent video monitoring includes:
S1, acquiring a preset monitoring video image, and generating a video interframe sequence diagram of the monitoring video image through a preset time attribute and a preset association attribute;
In the embodiment of the invention, the monitoring video image is captured according to the image captured by the camera, and the intelligent monitoring comprises an embedded video shooting function, and meanwhile, an intelligent recognition algorithm is integrated, so that the shot scene image can be rapidly analyzed, a corresponding response is made according to a set task, and an alarm signal can be generated to remind a user under appropriate conditions.
In detail, the preset monitoring video image may be acquired by a computer sentence (e.g., java sentence, python sentence, etc.) having a data capturing function.
Further, the monitoring video image is composed of images of one frame and one frame, so that the images of one frame and one frame are ordered according to a time relation, and abnormal data in a video sequence are analyzed.
In the embodiment of the invention, the time attribute means that each frame of image in the monitoring video image is represented by time, and for a monitoring video segment with N frames of video, continuous frames in the segment show inter-frame aggregation characteristic due to the time sequence relation. The correlation attribute is inter-frame aggregation characteristic, namely, a single frame j shows larger correlation with n adjacent frames before and after the single frame j, and the inter-frame correlation gradually decreases with the increase of the inter-frame distance. And for normal surveillance video segments, the aggregate correlation properties of adjacent frames should appear as uniform throughout, while for surveillance video segments where abnormal behavior or abnormal data occurs, the aggregate correlation properties of adjacent frames appear as cluttered independence.
In the embodiment of the present invention, referring to fig. 2, the generating a video inter-frame sequence chart of the monitoring video image according to the preset time attribute and the preset association attribute includes:
S21, generating a video frame node of the monitoring video image according to the time attribute;
S22, constructing an inter-frame relation between each video frame node according to the association attribute;
S23, associating the video frame nodes with the inter-frame relationship through a preset time sequence relationship to obtain a video inter-frame sequence diagram of the monitoring video image.
In detail, each frame image in the monitoring video image is taken as one video frame node according to the time point indicated by the time attribute, if the frame image corresponding to the time point 1 is the first frame, the frame image of the first frame is taken as one video frame node, and the frame image corresponding to the time point 2 is the second frame, and the frame image of the second frame is taken as another video frame node.
Specifically, the correlation attribute has larger similarity between the continuous frames of the normal video, so that the inter-frame change also has consistency and consistency, namely, the node edge relationship between each video frame node is determined through the correlation attribute, so that the inter-frame relationship between each video frame node is determined.
In the embodiment of the present invention, the constructing an inter-frame relationship between each video frame node according to the association attribute includes:
mapping each video frame node into a distance space according to the association attribute;
Determining node edges from node distances between the video frame nodes in the distance space, wherein the node edges are defined as:
；
Wherein, Represents the/>Video frame node and/>Node edge value of each video frame node,/>Represents the/>Video frame node and/>Node distance between individual video frame nodes,/>Representing node effective distance,/>Represent the firstVideo frame node and/>Node distance between individual video frame nodes,/>Representing real number set,/>Representing an exponential function;
And determining the inter-frame relationship between each video frame node according to the node edges.
In detail, the time relation of the video frame nodes is mapped into a distance space, and the relation between each video frame node is represented by distance, namely by a distance mapping functionConverting a point in time of a video frame node into a distance, wherein/>For/>Video frame node and/>The node distance between the nodes of the individual video frames,For/>Time points of the video frame nodes are converted into mapping functions of distances,/>For/>Time points of the video frame nodes are converted into mapping functions of distances,/>And mapping each video frame node to a distance space according to the association attribute for the node effective distance.
Specifically, the relevant node edges between the video frame nodes are determined through the node distance between each video frame node in the distance space, when the node edge is 1, the nodes are related, when the node edge is 0, the nodes are unrelated, and the weights between the nodes are determined according to the node distance. And then, according to the existence of the node edges, each video frame node is associated, so that the inter-frame relation between the video frame nodes is determined.
Further, the video frame nodes and the inter-frame relation are associated according to a preset time sequence relation, so that a video inter-frame sequence diagram of the monitoring video image can be obtained, whether abnormal behaviors or abnormal data exist in each frame or not is detected according to the video inter-frame sequence diagram, and the safety of the monitoring video is guaranteed.
S2, extracting video single-frame characteristics of the video inter-frame sequence diagram, and detecting abnormal data in the video single-frame characteristics by using a preset abnormal detection algorithm to obtain a video abnormal sequence;
In the embodiment of the invention, the continuous frames of the normal monitoring video have larger similarity, so that the inter-frame change also has continuity and consistency, the single frame characteristics of the video inter-frame sequence diagram are analyzed, and when a series of single frame characteristics are uniformly distributed, the video is more likely to be the normal video.
In detail, the video single-frame feature of the video inter-frame sequence chart can be extracted through a preset image feature extraction model (such as a convolutional neural network, a residual neural network and the like). The video single-frame feature refers to image pixel features and time features of a video inter-frame sequence chart of each frame.
Further, the data in the video single frame feature is analyzed, abnormal data in the video monitoring image is detected, a video abnormal sequence is formed, and the integrity and the accuracy of abnormal data detection are guaranteed.
In the embodiment of the invention, the anomaly detection algorithm is to compare the behavior or mode of special data with the behavior or model of normal data, determine the distance scale between the abnormal data and the normal data, and detect the abnormal data according to the distance scale.
In the embodiment of the present invention, referring to fig. 3, the detecting abnormal data in the video single frame feature by using a preset abnormal detection algorithm to obtain a video abnormal sequence includes:
s31, calculating the feature distance scale between each video frame and a preset normal frame one by one according to the video single frame features by using the anomaly detection algorithm;
S32, when the characteristic distance scale is larger than a preset characteristic distance threshold, taking the video frame corresponding to the characteristic distance scale as abnormal data;
S33, gathering video frames corresponding to the abnormal data to be the video abnormal sequence.
In detail, determining a distance value of each video frame according to the video single-frame features, namely determining the distance value of each video frame according to the gray value of the image features corresponding to the video single-frame features, and calculating feature distance scales between the distance value of each video frame and a preset distance threshold value of a normal frame one by one, namely subtracting the distance values between the two to obtain the feature distance scales.
Specifically, only when the feature distance scale is larger than a preset feature distance threshold, namely when the feature distance scale corresponding to the video frame is too large, abnormal operation or abnormal behavior of the video is indicated, the video frame corresponding to the feature distance scale is used as abnormal data, and all video frames corresponding to the abnormal data are collected to form a video abnormal sequence.
Further, the abnormal data in the video abnormal sequence is classified, so that the judgment of the abnormal data is more accurate in subdivision, and the detection accuracy of the abnormal data is ensured.
S3, carrying out classification recognition on the video abnormal sequence through a preset classification recognition model to obtain a video classification abnormal sequence, and determining an abnormal value of the video classification abnormal sequence according to a preset classification coefficient;
In the embodiment of the invention, the classification recognition model is constructed based on a decision tree model, and the abnormal category to which the abnormal data belong is classified to determine the importance level of the abnormal data. Wherein the decision tree is a tree structure in which each internal node represents a judgment on an attribute, each branch represents an output of a judgment result, and each leaf node represents a classification result.
In the embodiment of the present invention, before the video abnormal sequence is classified and identified by the preset classification and identification model, the method further includes:
Acquiring abnormal classification attributes of the monitoring video image;
Calculating the entropy value of each abnormal data in the preset video sample abnormal sequence according to the abnormal classification attribute by using the following entropy value calculation formula:
；
Wherein, Represents the/>Entropy of the abnormal data,/>Representing the amount of data in the abnormal sequence of video samples,/>Indicating that the anomaly data belongs to the/>Number of the abnormal classification attributes,/>Representing the classification quantity of the abnormal classification attribute,/>Representing a logarithmic function;
Selecting the abnormal data with the maximum entropy value as a branch node, marking the branch node as a classification attribute node, deleting the abnormal data corresponding to the branch node from the video sample abnormal sequence, and obtaining an updated video sample sequence;
When the classification attribute node exists, calculating an entropy value of each anomaly in the updated video sample sequence, and returning to the step of selecting the anomaly data with the maximum entropy value as a branch node until the classification attribute node does not exist;
And when the classification attribute node does not exist, generating the classification recognition model.
In detail, the abnormality classification attribute includes falsifying data, inserting abnormal data, deleting normal data, out-of-order data between frames, and the like. And carrying out classification training on the video sample sequence of the original abnormal data, calculating the entropy value of each abnormal data in the video sample abnormal sequence according to the abnormal classification attribute, selecting the abnormal data with the maximum entropy value as a branch node, and marking the abnormal classification attribute of the branch node.
Specifically, the entropy iterative computation is continuously carried out on the abnormal data until all classification attribute nodes exist in a preset decision tree, and the generated decision tree is used as the classification recognition model.
And further, carrying out abnormal classification on the video abnormal sequence by using the constructed classification recognition module, so as to determine the abnormal level of the video abnormal sequence, and carrying out corresponding early warning according to the abnormal level.
In the embodiment of the present invention, the classifying and identifying the video abnormal sequence through a preset classifying and identifying model to obtain a video classifying and identifying abnormal sequence includes:
Inputting the abnormal data in the video abnormal sequence one by one into the classification recognition model to perform classification judgment, and obtaining classification attributes corresponding to the abnormal data;
and collecting the abnormal data with the same classification attribute as the video classification abnormal sequence.
In detail, the abnormal data in the video abnormal sequence is input into the classification recognition model one by one for classification comparison, the abnormal data is compared one by one according to branch judgment in the classification recognition model, finally, the classification attribute corresponding to the abnormal data is obtained, and the abnormal data with the same classification attribute is collected to be used as the video classification abnormal sequence.
For example, if the video anomaly sequence includes video frame 1, video frame 2, and video frame 3 to video 7, the anomaly data corresponding to the video frame is classified according to the classification recognition model, and if the classification attribute corresponding to the video frame 1 and the video frame 3 is tampered data; the classification attribute corresponding to the video frame 2, the video frame 4 and the video frame 5 is inserted abnormal data; the classification attribute corresponding to the video frame 6 is deleting normal data; if the classification attribute corresponding to the video frame 7 is that the inter-frame data is disordered, the video frame 1 and the video frame 3 are used as a video classification abnormal sequence, and the video frame 2, the video frame 4 and the video frame 5 are used as a video classification abnormal sequence.
Further, in order to determine the anomaly importance of the anomaly data, an anomaly value of each video classification anomaly sequence needs to be determined, and then an anomaly level in the overall monitoring video image is determined according to the anomaly value.
In the embodiment of the present invention, the determining the abnormal value of the video classification abnormal sequence according to the preset classification coefficient includes:
counting the total quantity of abnormal data in the video classification abnormal sequence;
Calculating the outlier of the video classification outlier sequence according to the classification coefficient and the total number by using the outlier calculation function as follows:
；
Wherein, For/>Abnormal value of individual video classification abnormal sequence,/>For/>Classification coefficient of abnormal sequence of individual video classification,/>For/>Data quantity of abnormal sequence of individual video classification,/>Is the total number.
In detail, the number of the abnormal data in each video classification abnormal sequence is counted, the total number of the abnormal data in all the video classification abnormal sequences is counted, and then the abnormal value of each video classification abnormal sequence is calculated according to the total number of the abnormal data and a preset classification coefficient, namely, the number of the abnormal data in each video classification abnormal sequence is compared with the total data quantity to obtain a ratio, the abnormal duty ratio in the video classification abnormal sequence can be determined according to the ratio, and then the abnormal value of the video classification abnormal sequence is determined according to the abnormal duty ratio and the classification coefficient.
Specifically, the classification coefficient is determined according to the average value of entropy values corresponding to the abnormal data included in each video classification abnormal sequence. If the video classification abnormal sequence comprises a video frame 2, a video frame 4 and a video frame 5, entropy values corresponding to the video frame 2, the video frame 4 and the video frame 5 are obtained, an entropy value average value is calculated according to the entropy values, and the entropy value average value is used as a classification coefficient in the video classification abnormal sequence.
Further, corresponding total abnormal values of all abnormal data in the monitoring video image are determined according to the abnormal value of each video classification abnormal sequence, and a data safety early warning level is determined according to the total abnormal values.
S4, calculating a sequence anomaly threshold value of the video anomaly sequence according to the anomaly value and a preset dynamic weight through a preset dynamic weight algorithm;
In the embodiment of the invention, the sequence anomaly threshold value refers to the anomaly value of all video classification anomaly sequences and the anomaly value determined by the dynamic weight of the video anomaly sequences.
In the embodiment of the present invention, the calculating, by a preset dynamic weight algorithm, the sequence anomaly threshold value of the video anomaly sequence according to the anomaly value and the preset dynamic weight includes:
Dynamically monitoring the video abnormal sequence to obtain a dynamic video abnormal sequence;
Determining dynamic correction parameters of the dynamic video abnormal sequence;
Calculating a sequence anomaly threshold value of the video anomaly sequence according to the anomaly value, the dynamic weight and the dynamic correction parameter by the dynamic weight algorithm, wherein the dynamic weight algorithm is as follows:
；
Wherein, For the sequence anomaly threshold value,/>Classifying sequence numbers of abnormal sequences for video,/>For/>Dynamic weighting of individual video classification anomaly sequences,/>For/>Abnormal value of individual video classification abnormal sequence,/>For the dynamic correction parameters,/>As a logarithmic function.
In detail, the intelligent monitoring video is increased in real time, and abnormal data may occur in the process of increasing the video content, so that dynamic monitoring of the video abnormal sequence is required, and a dynamic video abnormal sequence can be obtained. The dynamic monitoring is to detect whether abnormal data are added or deleted in the video abnormal sequence, and determine a dynamic correction parameter according to the added normal data or the added abnormal data quantity, if the added abnormal data are 3, the dynamic correction parameter is 3; if the added abnormal data is 5 and the deleted normal data is 2, the dynamic correction coefficient is 7.
Specifically, the sequence anomaly threshold value of the video anomaly sequence is calculated through a dynamic weight algorithm, wherein the dynamic correction coefficient in the dynamic weight algorithm can realize real-time change of the video anomaly sequence, and the accuracy of calculating the sequence anomaly threshold value is ensured.
Further, an early warning value of the monitoring video image is determined according to the sequence abnormal critical value of the monitoring video image, and further, the data security of the monitoring video is accurately analyzed according to the early warning value.
And S5, calculating a safety early warning value of the monitoring video image according to the sequence abnormal critical value by using a preset safety early warning algorithm, and determining the data safety of the monitoring video image according to the safety early warning value.
In the embodiment of the invention, the safety early warning algorithm judges the abnormal value of the abnormal data under the condition of data safety, counts the number of sequences with the abnormal value higher than the sequence abnormal critical value, and further determines the safety early warning value of the current monitoring video.
In the embodiment of the present invention, the calculating the security pre-warning value of the monitoring video image according to the sequence anomaly threshold value by using a preset security pre-warning algorithm includes:
determining the number of sequences with abnormal values higher than the sequence abnormal critical value corresponding to the video classification abnormal sequence by utilizing the safety early warning algorithm;
calculating abnormal frequency of the monitoring video image according to the number of the sequences;
And determining a safety precaution value of the monitoring video image according to the abnormal frequency.
In detail, the number of sequences, corresponding to abnormal values higher than a sequence abnormal critical value, in the video classification abnormal sequences in the video abnormal sequences is counted through the safety early warning algorithm. If the outlier of the video classification abnormal sequence 1 is 0.6, the outlier of the video classification abnormal sequence 2 is 0.8, the outlier of the video classification abnormal sequence 3 is 0.3, and the sequence outlier threshold is 0.5, the number of the counted sequences is 2.
Specifically, comparing the number of sequences with the number of sequences of all video classification abnormal sequences, taking the obtained ratio as the abnormal frequency of the monitoring image, and taking the abnormal frequency percentage as the safety early warning value of the monitoring video image.
Further, the data security of the monitoring image is analyzed according to the security early warning value, so that the occurrence of unsafe conditions can be timely processed.
In the embodiment of the present invention, the determining the data security of the monitoring video image according to the security early warning value includes:
Determining the data security early warning level of the monitoring video image according to the security early warning value;
and triggering an alarm command for data security of the monitoring video image when the early warning level is higher than a preset security early warning level.
In detail, the early warning level of the safety early warning is divided into low level, medium level and high level, and the early warning level of the monitoring video image is determined according to the range of the safety early warning value. If the range of the safety early warning value is {0,45}, the early warning level is low; the range of the safety early warning value is {46,70}, and the early warning level is a middle level; the range of the safety early warning value is {71,100}, and the early warning level is high.
Specifically, when the early warning level is higher than a preset safety early warning level, an alarm command for data safety of the monitoring video image is triggered to ensure the data safety of the monitoring video.
According to the embodiment of the invention, the video inter-frame sequence diagram of the monitoring video image is generated through the time attribute and the associated attribute, so that abnormal data is detected according to the video inter-frame sequence diagram to obtain a video abnormal sequence, each frame of image in the monitoring video is accurately detected, and the integrity of abnormal data detection is realized; classifying according to the video abnormal sequence, further determining an abnormal value of the video classified abnormal sequence, and calculating a sequence abnormal critical value of the video abnormal sequence according to the abnormal value and the dynamic weight, so that accurate analysis of the data security of the complete monitoring video is facilitated; and determining the safety precaution value of the monitoring video image according to the sequence abnormal critical value, and further determining the data safety of the monitoring video image according to the safety precaution value, thereby being beneficial to improving the accuracy of the data safety precaution. Therefore, the data security early warning method and device based on intelligent video monitoring can solve the problem of low accuracy of data security early warning.
Fig. 4 is a functional block diagram of a data security early warning device based on intelligent video monitoring according to an embodiment of the present invention.
The data security early warning device 100 based on intelligent video monitoring can be installed in electronic equipment. According to the functions implemented, the data security pre-warning device 100 based on intelligent video monitoring may include a video inter-frame sequence diagram generating module 101, an abnormal data detecting module 102, an abnormal value calculating module 103, a sequence abnormal critical value calculating module 104 and a data security pre-warning module 105. The module of the invention, which may also be referred to as a unit, refers to a series of computer program segments, which are stored in the memory of the electronic device, capable of being executed by the processor of the electronic device and of performing a fixed function.
In the present embodiment, the functions concerning the respective modules/units are as follows:
The video inter-frame sequence diagram generating module 101 is configured to obtain a preset monitoring video image, and generate a video inter-frame sequence diagram of the monitoring video image according to a preset time attribute and a preset association attribute;
The abnormal data detection module 102 is configured to extract a video single frame feature of the video inter-frame sequence chart, detect abnormal data in the video single frame feature by using a preset abnormal detection algorithm, and obtain a video abnormal sequence;
the outlier calculating module 103 is configured to perform classification recognition on the video abnormal sequence through a preset classification recognition model to obtain a video classification abnormal sequence, and determine an outlier of the video classification abnormal sequence according to a preset classification coefficient;
The sequence anomaly threshold value calculation module 104 is configured to calculate a sequence anomaly threshold value of the video anomaly sequence according to the anomaly value and a preset dynamic weight through a preset dynamic weight algorithm;
The data security early-warning module 105 is configured to calculate a security early-warning value of the surveillance video image according to the sequence anomaly threshold value by using a preset security early-warning algorithm, and determine data security of the surveillance video image according to the security early-warning value.
In detail, each module in the data security pre-warning device 100 based on intelligent video monitoring in the embodiment of the present invention adopts the same technical means as the data security pre-warning method based on intelligent video monitoring described in fig. 1 to 3, and can generate the same technical effects, which is not repeated here.
In the several embodiments provided by the present invention, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is merely a logical function division, and there may be other manners of division when actually implemented.
The modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical units, may be located in one place, or may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional module in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units can be realized in a form of hardware or a form of hardware and a form of software functional modules.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof.
The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.
The embodiment of the application can acquire and process the related data based on the artificial intelligence technology. Wherein artificial intelligence (ARTIFICIAL INTELLIGENCE, AI) is the theory, method, technique, and application system that uses a digital computer or a digital computer-controlled machine to simulate, extend, and expand human intelligence, sense the environment, acquire knowledge, and use knowledge to obtain optimal results.
Furthermore, it is evident that the word "comprising" does not exclude other elements or steps, and that the singular does not exclude a plurality. Multiple units or means as set forth in the system embodiments may also be implemented by one unit or means in software or hardware. The terms first, second, etc. are used to denote a name, but not any particular order.
Finally, it should be noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made to the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention.
Claims (10)
1. The data security early warning method based on intelligent video monitoring is characterized by comprising the following steps:
S1, acquiring a preset monitoring video image, and generating a video interframe sequence diagram of the monitoring video image through a preset time attribute and a preset association attribute;
s2, extracting video single-frame characteristics of the video inter-frame sequence diagram, and detecting abnormal data in the video single-frame characteristics by using a preset abnormal detection algorithm to obtain a video abnormal sequence;
S3, carrying out classification recognition on the video abnormal sequence through a preset classification recognition model to obtain a video classification abnormal sequence, and determining an abnormal value of the video classification abnormal sequence according to a preset classification coefficient;
s4, calculating a sequence anomaly threshold value of the video anomaly sequence according to the anomaly value and a preset dynamic weight through a preset dynamic weight algorithm, wherein the calculating the sequence anomaly threshold value of the video anomaly sequence according to the anomaly value and the preset dynamic weight through the preset dynamic weight algorithm comprises the following steps:
s41, dynamically monitoring the video abnormal sequence to obtain a dynamic video abnormal sequence;
S42, determining dynamic correction parameters of the dynamic video abnormal sequence;
s43, calculating a sequence anomaly threshold value of the video anomaly sequence according to the anomaly value, the dynamic weight and the dynamic correction parameter through the dynamic weight algorithm, wherein the dynamic weight algorithm is as follows:
；
Wherein, For the sequence anomaly threshold value,/>Classifying sequence numbers of abnormal sequences for video,/>For/>Dynamic weighting of individual video classification anomaly sequences,/>For/>Abnormal value of individual video classification abnormal sequence,/>For the dynamic correction parameters,/>As a logarithmic function;
And S5, calculating a safety early warning value of the monitoring video image according to the sequence abnormal critical value by using a preset safety early warning algorithm, and determining the data safety of the monitoring video image according to the safety early warning value.
2. The method for data security pre-warning based on intelligent video monitoring as claimed in claim 1, wherein the generating the video inter-frame sequence diagram of the monitoring video image by the preset time attribute and the preset association attribute comprises:
generating a video frame node of the monitoring video image according to the time attribute;
Constructing an inter-frame relation between each video frame node according to the association attribute;
And associating the video frame nodes with the inter-frame relationship through a preset time sequence relationship to obtain a video inter-frame sequence diagram of the monitoring video image.
3. The method for data security pre-warning based on intelligent video monitoring according to claim 2, wherein said constructing an inter-frame relationship between each of said video frame nodes according to said association attribute comprises:
mapping each video frame node into a distance space according to the association attribute;
Determining node edges from node distances between the video frame nodes in the distance space, wherein the node edges are defined as:
；
Wherein, Represents the/>Video frame node and/>Node edge value of each video frame node,/>Represents the/>Video frame node and/>Node distance between individual video frame nodes,/>Representing node effective distance,/>Represents the/>Video frame node and/>Node distance between individual video frame nodes,/>Representing real number set,/>Representing an exponential function;
And determining the inter-frame relationship between each video frame node according to the node edges.
4. The method for data security early warning based on intelligent video monitoring according to claim 1, wherein the detecting abnormal data in the video single frame feature by using a preset abnormality detection algorithm to obtain a video abnormal sequence comprises:
Calculating the feature distance scale between each video frame and a preset normal frame one by one according to the video single frame features by using the anomaly detection algorithm;
When the characteristic distance scale is larger than a preset characteristic distance threshold, taking the video frame corresponding to the characteristic distance scale as abnormal data;
and collecting video frames corresponding to the abnormal data as the video abnormal sequence.
5. The method for data security early warning based on intelligent video monitoring according to claim 1, wherein before the video abnormal sequence is classified and identified by a preset classification and identification model to obtain a video classified abnormal sequence, the method further comprises:
Acquiring abnormal classification attributes of the monitoring video image;
Calculating the entropy value of each abnormal data in the preset video sample abnormal sequence according to the abnormal classification attribute by using the following entropy value calculation formula:
；
Wherein, Represents the/>Entropy of the abnormal data,/>Representing the amount of data in the abnormal sequence of video samples,/>Indicating that the anomaly data belongs to the/>Number of the abnormal classification attributes,/>Representing the classification quantity of the abnormal classification attribute,/>Representing a logarithmic function;
Selecting the abnormal data with the maximum entropy value as a branch node, marking the branch node as a classification attribute node, deleting the abnormal data corresponding to the branch node from the video sample abnormal sequence, and obtaining an updated video sample sequence;
When the classification attribute node exists, calculating an entropy value of each anomaly in the updated video sample sequence, and returning to the step of selecting the anomaly data with the maximum entropy value as a branch node until the classification attribute node does not exist;
And when the classification attribute node does not exist, generating the classification recognition model.
6. The method for data security early warning based on intelligent video monitoring according to claim 1, wherein the step of classifying and identifying the video abnormal sequence through a preset classification model to obtain a video classification abnormal sequence comprises the following steps:
Inputting the abnormal data in the video abnormal sequence one by one into the classification recognition model to perform classification judgment, and obtaining classification attributes corresponding to the abnormal data;
and collecting the abnormal data with the same classification attribute as the video classification abnormal sequence.
7. The method for data security pre-warning based on intelligent video monitoring according to claim 1, wherein the determining the abnormal value of the video classification abnormal sequence according to the preset classification coefficient comprises:
counting the total quantity of abnormal data in the video classification abnormal sequence;
Calculating the outlier of the video classification outlier sequence according to the classification coefficient and the total number by using the outlier calculation function as follows:
；
Wherein, For/>Abnormal value of individual video classification abnormal sequence,/>For/>Classification coefficient of abnormal sequence of individual video classification,/>For/>Data quantity of abnormal sequence of individual video classification,/>Is the total number.
8. The intelligent video monitoring-based data security pre-warning method according to any one of claims 1 to 7, wherein the calculating the security pre-warning value of the monitoring video image according to the sequence anomaly threshold value by using a preset security pre-warning algorithm comprises:
determining the number of sequences with abnormal values higher than the sequence abnormal critical value corresponding to the video classification abnormal sequence by utilizing the safety early warning algorithm;
calculating abnormal frequency of the monitoring video image according to the number of the sequences;
And determining a safety precaution value of the monitoring video image according to the abnormal frequency.
9. The method for data security early warning based on intelligent video monitoring according to claim 1, wherein the determining the data security of the monitoring video image according to the security early warning value comprises:
Determining the data security early warning level of the monitoring video image according to the security early warning value;
and triggering an alarm command for data security of the monitoring video image when the early warning level is higher than a preset security early warning level.
10. A data security early warning device based on intelligent video monitoring, for implementing the data security early warning method based on intelligent video monitoring as set forth in claim 1, the device comprising:
The video inter-frame sequence diagram generation module is used for acquiring a preset monitoring video image and generating a video inter-frame sequence diagram of the monitoring video image through a preset time attribute and a preset association attribute;
The abnormal data detection module is used for extracting video single frame characteristics of the video inter-frame sequence diagram, and detecting abnormal data in the video single frame characteristics by using a preset abnormal detection algorithm to obtain a video abnormal sequence;
The abnormal value calculation module is used for carrying out classification recognition on the video abnormal sequence through a preset classification recognition model to obtain a video classification abnormal sequence, and determining an abnormal value of the video classification abnormal sequence according to a preset classification coefficient;
The sequence anomaly threshold value calculation module is used for calculating the sequence anomaly threshold value of the video anomaly sequence according to the anomaly value and the preset dynamic weight through a preset dynamic weight algorithm;
And the data safety early warning module is used for calculating the safety early warning value of the monitoring video image according to the sequence abnormal critical value by utilizing a preset safety early warning algorithm and determining the data safety of the monitoring video image according to the safety early warning value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410294857.2A CN117911930B (en) | 2024-03-15 | 2024-03-15 | Data security early warning method and device based on intelligent video monitoring |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410294857.2A CN117911930B (en) | 2024-03-15 | 2024-03-15 | Data security early warning method and device based on intelligent video monitoring |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117911930A CN117911930A (en) | 2024-04-19 |
| CN117911930B true CN117911930B (en) | 2024-06-04 |
Family
ID=90695055
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410294857.2A Active CN117911930B (en) | 2024-03-15 | 2024-03-15 | Data security early warning method and device based on intelligent video monitoring |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117911930B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103390278A (en) * | 2013-07-23 | 2013-11-13 | 中国科学技术大学 | Detecting system for video aberrant behavior |
| CN103929592A (en) * | 2014-04-22 | 2014-07-16 | 杭州道联电子技术有限公司 | All-dimensional intelligent monitoring equipment and method |
| CN111931587A (en) * | 2020-07-15 | 2020-11-13 | 重庆邮电大学 | Video anomaly detection method based on interpretable space-time self-encoder |
| CN114882251A (en) * | 2022-05-09 | 2022-08-09 | 深圳市中电网络技术有限公司 | Internet of things big data intelligent video security monitoring method and device |
| CN117197713A (en) * | 2023-09-08 | 2023-12-08 | 南京敏锐科技有限公司 | Extraction method based on digital video monitoring system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102711881B1 (en) * | 2022-02-23 | 2024-10-02 | 세종대학교산학협력단 | Method and Apparatus for Multi-Frame Prediction Error-based Video Anomaly Detection |
-
2024
- 2024-03-15 CN CN202410294857.2A patent/CN117911930B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103390278A (en) * | 2013-07-23 | 2013-11-13 | 中国科学技术大学 | Detecting system for video aberrant behavior |
| CN103929592A (en) * | 2014-04-22 | 2014-07-16 | 杭州道联电子技术有限公司 | All-dimensional intelligent monitoring equipment and method |
| CN111931587A (en) * | 2020-07-15 | 2020-11-13 | 重庆邮电大学 | Video anomaly detection method based on interpretable space-time self-encoder |
| CN114882251A (en) * | 2022-05-09 | 2022-08-09 | 深圳市中电网络技术有限公司 | Internet of things big data intelligent video security monitoring method and device |
| CN117197713A (en) * | 2023-09-08 | 2023-12-08 | 南京敏锐科技有限公司 | Extraction method based on digital video monitoring system |
Non-Patent Citations (1)
| Title |
|---|
| Anomaly Detection in Videos for Video Surveillance Applications using Neural Networks;Ruben J Franklin 等;2020 Fourth International Conference on Inventive Systems and Control (ICISC);20200110;632-637 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117911930A (en) | 2024-04-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113792453B (en) | Digital twinning-based partial discharge monitoring system, method and device | |
| CN109977895B (en) | Wild animal video target detection method based on multi-feature map fusion | |
| CN111723773B (en) | Method and device for detecting carryover, electronic equipment and readable storage medium | |
| CN111176953B (en) | Abnormality detection and model training method, computer equipment and storage medium | |
| CN113270197A (en) | Health prediction method, system and storage medium based on artificial intelligence | |
| CN111898581A (en) | Animal detection method, device, electronic equipment and readable storage medium | |
| CN111291817A (en) | Image recognition method and device, electronic equipment and computer readable medium | |
| CN111818198A (en) | Domain name detection method, domain name detection device, equipment and medium | |
| CN115660262B (en) | Engineering intelligent quality inspection method, system and medium based on database application | |
| CN109117771B (en) | System and method for detecting violence events in image based on anchor nodes | |
| CN116453056A (en) | Target detection model construction method and transformer substation foreign matter intrusion detection method | |
| CN113076963A (en) | Image recognition method and device and computer readable storage medium | |
| CN115545103A (en) | Abnormal data identification method, label identification method and abnormal data identification device | |
| CN117765348A (en) | Target detection model deployment method, target detection method and electronic equipment | |
| CN117911930B (en) | Data security early warning method and device based on intelligent video monitoring | |
| Wang et al. | Improved faster-RCNN algorithm for mask wearing detection | |
| CN116168213A (en) | People flow data identification method and training method of people flow data identification model | |
| CN113743293B (en) | Fall behavior detection method and device, electronic equipment and storage medium | |
| CN114172856B (en) | Message automatic replying method, device, equipment and storage medium | |
| CN116775400A (en) | Log analysis method, device, equipment and storage medium | |
| CN111209567B (en) | Method and device for judging perceptibility of improving robustness of detection model | |
| CN109670470B (en) | Pedestrian relationship identification method, device and system and electronic equipment | |
| CN115514621B (en) | Fault monitoring method, electronic device and storage medium | |
| CN110728615B (en) | Steganalysis method based on sequential hypothesis testing, terminal device and storage medium | |
| CN118447459B (en) | Landslide accumulation monitoring method and system based on deep learning |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |