Methods for video image data quality detection, evaluation and optimization management

Abstract

The present invention discloses a method for video image data quality detection, evaluation and optimization management. The method for video image data quality detection, evaluation and optimization management includes a video image data access process, a video image data detection process and a video image data management process. Compared with the prior art, the technical solution of the present invention has the advantages of systematic and comprehensive detection, evaluation, assessment and calibration optimization of video images, which is not limited to one dimension or one type of video image data; it provides automatic management and manual management solutions, covering the management of all view data quality issues; and it forms a closed loop from access management data to output after management.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method for detecting, evaluating and optimizing video image data quality. Background Art

[0002] In recent years, video surveillance has been widely used in urban management, intelligent security monitoring and other fields. A large number of front-end monitoring equipment have been built, producing massive video image data. These video image data include view device data, video stream data, image data and ZDR trajectory data. From the data type point of view, it includes structured data, unstructured data and semi-structured data.

[0003] Due to the lack of supervision in the information collection process of front-end equipment, the lack of a continuous and effective operation and maintenance guarantee system, the numerous standard deviations of image analysis vendors, and the differences in ZDR trajectory comparison and early warning algorithms and processes in various regions, there are serious quality problems in video image data, which seriously affects the in-depth application of image data in urban management, security monitoring and policing.

[0004] At present, the solutions to the quality problem of video image data mainly solve the quality problem from the operation and maintenance perspective or the business perspective, and often only solve one type of data or one aspect of the problem; for example, the video operation and maintenance technical solution is to detect the video stream data from the aspects of video stream online status, network status and video quality; the view library system mainly performs quality detection on image data from the single data dimension of image data, and the equipment file type mainly performs quality detection from the single dimension of equipment information.

[0005] The shortcomings of the above solutions are:

[0006] 1. Starting from a single type of data or a single business, it can only solve the quality inspection problem of one type of data;

[0007] 2. Mainly solve quality inspection problems, without effective calibration optimization and other governance capabilities;

[0008] 3. Lack of systematic video image data evaluation and assessment system. Summary of the invention

[0009] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a method for video image data quality detection, evaluation and optimization management, so as to systematically and comprehensively detect, evaluate, assess and calibrate video images, and is not limited to one dimension or one type of video image data.

[0010] The technical solution of the present invention is as follows:

[0011] Methods for video image data quality detection, evaluation and optimization management, including video image data access process, video image data detection process and video image data management process;

[0012] The steps of video image data access process are as follows:

[0013] The access to different video image data types is realized through the view data standard access module, and the video image data is stored in the access data pool after access;

[0014] Video image data sharing is provided from the shared data pool through the view data standard sharing module; when it is detected that video image data needs to be shared externally, it is pushed externally through the GA / T 1400 protocol;

[0015] The data detection process steps are as follows:

[0016] When video image data quality detection is required, the data sampling module will extract video image data from the access data pool according to the data sampling rules and store the sampled data into the database to be detected;

[0017] The sampling quality detection module reads the prepared sampling data and performs sampling quality detection according to the quality requirements of the relevant standards of the video data, including the GB / T 28181 protocol, the GA / T 1400 protocol requirements and the specific index requirements of the user unit for the data, according to different video image data types. After the sampling quality detection module completes the quality detection, it pushes the quality detection results to the data evaluation module;

[0018] The data evaluation module statistically sorts out the sampling test results, evaluates and calculates the compliance rate of the corresponding quality index items of the video image data, and outputs an evaluation report;

[0019] The data assessment module calculates the video image data quality assessment results of the current unit and the quality assessment results of subordinate units according to the corresponding relationship between the weights and scores of the video image data indicators formulated by the user unit, and outputs the assessment ranking information;

[0020] The steps of the data governance process are as follows:

[0021] Obtain the full amount of video image data accessed by the full amount of data quality detection module from the access data pool, and push it to the full amount of data quality detection module;

[0022] After receiving the video image data, the full data quality detection module performs full detection according to the data quality requirements of the GB / T 28181 protocol and the GA / T 1400 protocol and the specific index requirements of the user unit for the data, and pushes the detected video image data to the automatic calibration module in real time;

[0023] Determine whether the quality is qualified. If it is an unqualified video image, determine whether automatic treatment is required based on the type of video image problem. If automatic treatment is required, the video image will be handed over to the automatic calibration module for treatment; the video images that cannot be treated will be stored in the backup library for backup;

[0024] When the video image is judged to be unable to be automatically managed after detection, the manual calibration optimization module is used to generate a management work order for manual management;

[0025] The result of the management of the video image in the above steps is pushed to the qualified data sharing push module, and the qualified data sharing push module then pushes the video image to the shared data pool to complete the management operation.

[0026] The access standard requirements for the above video image data types are as follows: the basic data of the equipment shall be connected through the GB / T 28181 protocol, the GA / T 1400 protocol or the public security one-machine-one-file standard protocol; the video stream data shall be connected through the GB / T 28181 protocol; the image data shall be connected through the GA / T1400 protocol; and the ZDR trajectory data shall be connected through the GA / T 1400 protocol.

[0027] The sharing of the above video image data includes equipment basic data, image data and ZDR trajectory data.

[0028] The basis for the above-mentioned sampling data quality inspection is as follows: the equipment basic data refers to the relevant requirements for basic attributes in GB / T 28181 and GA / T 1400 and the specific rules and requirements of the user unit for the equipment basic data; the video stream data refers to the requirements for video stream data in GB / T 28181 and the specific rules and requirements of the user unit for video stream data; the image data refers to the basic requirements for image data in GA / T 1400 and the specific rules and requirements of the user unit for image data; the ZDR trajectory data refers to the basic requirements for image data in GA / T 1400 and the specific rules and requirements of the user unit for ZDR trajectory data.

[0029] The above-mentioned video image data management requirements are divided into: basic device data management: automatic management will perform cleaning, completion calibration and optimization operations on video image data; video stream data management: real-time video streams and historical usage streams do not support automatic management, and the device clock and device subtitle standards can be automatically modified and calibrated in batches using automatic management; image type data management: separate unqualified image data, and calibrate inaccurate image data by calling the current manufacturer's algorithm service or adding a new algorithm service; ZDR trajectory data: record problematic ZDR trajectory data and delete the source data from the original ZDR trajectory data set.

[0030] The above-mentioned contents that cannot be manually managed include basic equipment data and video stream data; the basic equipment data specifically refers to errors in location information in the equipment data, errors in or omissions in the attributes of whether the place where the equipment belongs is an important place; the video stream data specifically refers to problems such as the real-time video stream cannot be accessed and the historical video stream cannot be accessed.

[0031] Beneficial effects of the present invention: Compared with the prior art, the technical solution of the present invention has the following advantages:

[0032] 1. Systematically and comprehensively detect, evaluate, assess and calibrate video images, not limited to one dimension or one type of video image data;

[0033] 2. Provide automatic and manual governance solutions to cover the governance of all view data quality issues;

[0034] 3. Form a closed loop from access management data to post-management output. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 Flowchart of the method for video image data quality detection, evaluation and optimization management. DETAILED DESCRIPTION

[0036] In order to better illustrate the present invention, further description is now given in conjunction with embodiments and drawings.

[0037] The method of video image data quality detection, evaluation and optimization management includes a video image data access process, a video image data detection process and a video image data management process.

[0038] The video image data access process includes the following steps:

[0039] S11: Access to different types of video image data is achieved through the view data standard access module; the standard requirements for video image data type access are as follows: (1) Equipment basic data: connected through the GB / T 28181 protocol, GA / T 1400 protocol or the public security one machine one file standard protocol; (2) Video stream data: connected through the GB / T28181 protocol; (3) Image data: connected through the GA / T 1400 protocol; (4) ZDR trajectory data: connected through the GA / T 1400 protocol; After access, the video image data is stored in the access data pool;

[0040] S12: Provide video image data sharing from the shared data pool through the view data standard sharing module; the shared video image data includes basic equipment data, image data and ZDR trajectory data; and monitor the shared data pool in real time. When it is detected that there is video image data that needs to be shared externally, it is pushed externally through the GA / T 1400 protocol.

[0041] The video image data detection process includes the following steps:

[0042] S21: When video image data quality detection is required, the data sampling module will extract device basic data, video stream data (device information association), image data and ZDR trajectory data from the access data pool according to the data sampling rules, and store the sampled data into the database to be detected;

[0043] S22: The sampling quality detection module reads the prepared sampling data and performs sampling quality detection according to the quality requirements of the relevant standards of the view data, including the requirements of the GB / T 28181 protocol, the GA / T 1400 protocol and the specific index requirements of the user unit for the data, according to different video image data types. After the sampling quality detection module completes the quality detection, it pushes the quality detection results to the data evaluation module; the basis for quality detection is: (1) Equipment basic data: the relevant requirements for basic attributes in GB / T 28181 and GA / T1400 and the specific rules and requirements of the user unit for equipment basic data; (2) Video stream data: the requirements for video stream data in GB / T28181 and the specific rules and requirements of the user unit for video stream data; (3) Image data: the basic requirements for image data in GA / T 1400 and the specific rules and requirements of the user unit for image data; (4) ZDR trajectory data: GA / T The basic requirements for image data in 1400 and the specific rules and requirements for ZDR trajectory data by the user unit; after the sampling quality detection module completes the quality detection, the quality detection results are pushed to the data evaluation module;

[0044] S23: The data evaluation module statistically sorts out the sampling detection results, evaluates and calculates the compliance rate of the quality index items corresponding to the video image data, and outputs an evaluation report;

[0045] S24: The data assessment module calculates the video image data quality assessment score of the current unit and the quality assessment score of the subordinate units according to the corresponding relationship between the weights and scores of the video image data index items formulated by the user unit, and outputs the assessment ranking information;

[0046] The data governance process includes the following steps:

[0047] S31: Obtain the full amount of video image data accessed by the full amount of data quality detection module from the access data pool, and push it to the full amount of data quality detection module;

[0048] S32: After receiving the video image data, the full data quality detection module performs full detection according to the data quality requirements of the GB / T 28181 protocol and the GA / T1400 protocol and the specific index requirements of the user unit for the data, and pushes the detected video image data to the automatic calibration module in real time; the detection rules are as follows: (1) Equipment basic data: the relevant requirements for basic attributes in GB / T 28181 and GA / T 1400 and the specific rules and requirements of the user unit for equipment basic data; (2) Video stream data: the requirements for video stream data in GB / T 28181 and the specific rules and requirements of the user unit for video stream data; (3) Image data: the basic requirements for image data in GA / T 1400 and the specific rules and requirements of the user unit for image data; (4) ZDR trajectory data: the basic requirements for image data in GA / T 1400 and the specific rules and requirements of the user unit for ZDR trajectory data;

[0049] S33: Determine whether the quality is qualified. If not, determine whether automatic management is required based on the type of video image data problem. If automatic management is required, the video image data will be handed over to the automatic calibration optimization module for management; the four types of data management requirements for video image data are as follows: (1) Basic device data: Automatic management will perform cleaning, completion and calibration optimization operations on the video image data; (2) Video stream data: If real-time video and historical usage streams do not support automatic management, the device clock and device subtitle standards can be automatically modified and calibrated in batches using automatic management; (3) Image data: Separate unqualified image data, and calibrate inaccurate image data by calling the current manufacturer's algorithm service or a newly added algorithm service; (4) ZDR trajectory data: Record the problematic ZDR trajectory data and delete the source data from the original ZDR trajectory data. For video image data that cannot be managed, the automatic calibration optimization module will store the video image data in the backup library for backup;

[0050] S34: When the video image data is judged to be unable to be automatically managed after quality inspection, the manual calibration optimization module is used to generate a management work order for manual management. The contents that cannot be manually managed include: (1) Equipment basic data: The location information in the equipment basic data is wrong, and the attribute of whether the place where the equipment belongs is an important place is wrong or missing; (2) Video stream data: The real-time video stream cannot be accessed, and the historical video stream cannot be accessed.

[0051] S35: The result of the management of the video image data in the above steps S33 and S34 is pushed to the qualified data sharing push module, and the qualified data sharing push module pushes the video image data to the data sharing pool to complete the management operation.

[0052] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modification, equivalent substitution or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting, evaluating and optimizing video image data quality, characterized in that: Including video image data access process, video image data detection process and video image data management process; The steps of video image data access process are as follows: The access to different video image data types is realized through the view data standard access module, and the video image data is stored in the access data pool after access; Video image data sharing is provided from the shared data pool through the view data standard sharing module; when it is detected that video image data needs to be shared externally, it is pushed externally through the GA / T 1400 protocol; The data detection process steps are as follows: When video image data quality detection is required, the data sampling module will extract video image data from the access data pool according to the data sampling rules and store the sampled data into the database to be detected; The sampling quality detection module reads the prepared sampling data and performs sampling quality detection according to the quality requirements of the relevant standards of the video data, including the GB / T 28181 protocol, the GA / T 1400 protocol requirements and the specific index requirements of the user unit for the data, according to different video image data types. After the sampling quality detection module completes the quality detection, it pushes the quality detection results to the data evaluation module; The data evaluation module statistically sorts out the sampling test results, evaluates and calculates the compliance rate of the corresponding quality index items of the video image data, and outputs an evaluation report; The data assessment module calculates the video image data quality assessment results of the current unit and the quality assessment results of subordinate units according to the corresponding relationship between the weights and scores of the video image data indicators formulated by the user unit, and outputs the assessment ranking information; The steps of the data governance process are as follows: Obtain the full amount of video image data accessed by the full amount of data quality detection module from the access data pool, and push it to the full amount of data quality detection module; After receiving the video image data, the full data quality detection module performs full detection according to the data quality requirements of the GB / T 28181 protocol and the GA / T 1400 protocol and the specific index requirements of the user unit for the data, and pushes the detected video image data to the automatic calibration module in real time; Determine whether the quality is qualified. If it is an unqualified video image, determine whether automatic treatment is required based on the type of video image problem. If automatic treatment is required, the video image will be handed over to the automatic calibration module for treatment; the video images that cannot be treated will be stored in the backup library for backup; When the video image is judged to be unable to be automatically managed after detection, the manual calibration optimization module is used to generate a management work order for manual management; The result of the management of the video image in the above steps is pushed to the qualified data sharing push module, and the qualified data sharing push module then pushes the video image to the shared data pool to complete the management operation.

2. The method for video image data quality detection, evaluation and optimization management according to claim 1 is characterized by: The access standard requirements for the video image data type are as follows: the basic data of the device is connected through the GB / T 28181 protocol, the GA / T 1400 protocol or the public security one-machine-one-file standard protocol; the video stream data is connected through the GB / T 28181 protocol; the image data is connected through the GA / T1400 protocol; and the ZDR trajectory data is connected through the GA / T 1400 protocol.

3. The method for video image data quality detection, evaluation and optimization management according to claim 1 is characterized by: The sharing of the video image data includes equipment basic data, image data and ZDR trajectory data.

4. The method for video image data quality detection, evaluation and optimization management according to claim 1 is characterized by: The basis for the sampling data quality inspection is as follows: the basic equipment data refers to the relevant requirements for basic attributes in GB / T 28181 and GA / T 1400 and the specific rules and requirements of the user unit for the basic equipment data; the video stream data refers to the requirements for video stream data in GB / T 28181 and the specific rules and requirements of the user unit for video stream data; the image data refers to the basic requirements for image data in GA / T 1400 and the specific rules and requirements of the user unit for image data; the ZDR trajectory data refers to the basic requirements for image data in GA / T 1400 and the specific rules and requirements of the user unit for ZDR trajectory data.

5. The method for video image data quality detection, evaluation and optimization management according to claim 1 is characterized by: The requirements for video image data management are divided into: basic device data management: automatic management will perform cleaning, completion calibration and optimization operations on video image data; video stream data management: real-time video streams and historical usage streams do not support automatic management, and device clocks and device subtitle standards can be automatically modified and calibrated in batches using automatic management; image type data management: separate unqualified image data, and calibrate inaccurate image data by calling the current manufacturer's algorithm service or adding a new algorithm service; ZDR trajectory data: record problematic ZDR trajectory data and delete the source data from the original ZDR trajectory data set.

6. The method for video image data quality detection, evaluation and optimization management according to claim 1 is characterized by: The content that cannot be manually managed includes basic equipment data and video stream data; the basic equipment data specifically refers to errors in location information in the equipment data and errors or omissions in the attributes of whether the location of the equipment is an important location; the video stream data specifically refers to problems such as the inability to access real-time video streams and historical video streams.

Images