A method for optimizing process videos based on high-precision positioning

Abstract

The present invention proposes a method for optimizing process videos based on high-precision positioning in the fields of discrete manufacturing enterprises and Industry 4.0 intelligent manufacturing. This method combines video linkage and high-precision positioning technology to obtain the process video information of corresponding production equipment by calculating the position information of the production equipment. Regarding the key issues such as whether the collected videos can truly reflect the actual situation of each process and whether the videos are complete, the present invention introduces the index of video integrity to quantify the integrity degree of process videos. When the video integrity is lower than the video integrity threshold set by the system, solutions such as adjusting the video area or dynamically adjusting the video recording time can be adopted respectively to supplement the missing video information specifically or delete the redundant video information specifically, effectively improving the matching degree and integrity between the process videos and the actual production situation. This can not only control the production quality of each process point on the production line and improve production efficiency, but also contribute to the big data analysis and mining of Industry 4.0 intelligent manufacturing process videos, bringing huge direct and indirect economic values.

Description

Technical Field

[0001] The present invention relates to the field of video linkage and high-precision positioning technology, and has application prospects in discrete intelligent manufacturing enterprises. Background Art

[0002] In discrete intelligent manufacturing enterprises, how to accurately control the quality of production equipment in each process and fully realize visualization, digitization, informatization, refinement and intelligent management is a pain point problem that needs to be solved urgently. At present, the application of high-precision positioning and video linkage technology in Industry 4.0 and intelligent manufacturing is gradually becoming a development trend. For example, based on ultra-wideband UWB positioning technology or other sensor positioning technology, the smart factory system can obtain the video information of the corresponding production equipment by calculating the location information of the production equipment. However, in the complex factory assembly line production operation, there are a large number of mechanical equipment, mechanical arms, construction vehicles, workers, etc., and the complex environment will seriously affect the reliability and accuracy of positioning, affect the positioning accuracy, and then affect the integrity of the process video. Therefore, the quality of these discrete videos, whether they can truly reflect the actual situation in each process, and whether the video is complete, there is still a lack of effective solutions for these more valuable big data analysis problems.

[0003] Aiming at the fields of Industry 4.0 smart manufacturing and smart factories, this paper proposes a process video optimization method based on high-precision positioning. By introducing the video integrity index, the integrity of the process video can be quantified. At the same time, for how to improve the integrity of the process video, an effective solution based on adjusting the video area and dynamically adjusting the video recording time is proposed. Summary of the invention

[0004] The present invention discloses a process video optimization method based on high-precision positioning, and the method content is as follows.

[0005] Step 1: Build a high-precision electronic map of the smart factory production workshop, import the high-precision electronic map into the smart manufacturing factory management system platform, and draw a polygonal electronic fence on the high-precision electronic map according to the actual physical space area of ​​the nth process or several processes to represent the spatial area of ​​the nth process or several processes.

[0006] Step 2: Bind the kth camera deployed on the production line with the number of the nth process or some processes, and map the recording area of ​​the kth camera with the spatial area of ​​the nth process or some processes.

[0007] Step 3: When any moving entity carrying a positioning device enters the spatial area of the nth process or some processes corresponding to the kth camera, the intelligent manufacturing plant management system platform calculates the precise position information of the moving entity carrying the positioning device and compares it with the polygon electronic fence representing the spatial area of the nth process or some processes on the high-precision electronic map; if the position information of the moving entity carrying the positioning device has entered the polygon electronic fence of the spatial area of the nth process or some processes on the high-precision electronic map, the intelligent manufacturing plant management system platform will schedule the kth camera to start recording and synchronously storing the video of the corresponding process until any of the moving entities leaves the spatial area of the certain process or some processes corresponding to the kth camera.

[0008] Step 4: Retrieve the videos of a certain number of moving entities entering the spatial area of the nth process (n = 1, 2, …, N) or some processes corresponding to the kth camera (k = 1, 2, …, K), and calculate the video integrity of the nth process or some processes corresponding to the kth camera.

[0009] The video integrity of the nth process or some processes is an indicator of the integrity and accuracy of a certain number of process videos, and can be expressed as

[0010]

[0011] In the formula, η n is the video integrity of the nth process or some processes, M is the total number of videos of the nth process or some processes retrieved, and δ m,n is the video integrity identifier of the mth (m = 1, 2, …, M) video of the nth process or some processes; the process video must be a complete, clear, and accurate semi-structured video, usually including three stages, namely the stage when the moving entity enters the process, the stage when the moving entity is processed and produced, and the stage when the moving entity leaves the process.

[0012] If the mth video of the nth process or some processes completely matches the actual process, then record δ m,n = 1; if the mth video of the nth process or some processes does not match the actual process, including but not limited to the lack of the stage when the moving entity enters the process, and / or the lack of the stage when the moving entity is processed and produced, and / or the lack of the stage when the moving entity leaves the process, and / or there is a large amount of irrelevant redundant parts in the video, then record δ m,n = 0.

[0013] Step 5: If the video integrity of the nth process or some processes corresponding to the kth camera is lower than the video integrity threshold set by the system, it is necessary to optimize the recording strategy for the video of the nth process or some processes corresponding to the kth camera, so as to make the video integrity of the nth process or some processes corresponding to the kth camera better than the video integrity threshold set by the intelligent manufacturing plant management system platform.

[0014] Among them, the video integrity threshold is the threshold value of the process video integrity set according to different moving entities, denoted as Γ, and its value range is between [0,1]; generally, for moving entities with high technology maturity, the video integrity threshold Γ can be set relatively small, preferably set to 60% - 70%, which can relatively save the storage space of the process video and does not affect the system's quality control of the moving entity in each process link; for moving entities with relatively low technology maturity, the video integrity threshold Γ can be set relatively small, preferably set to 85% - 95%, and it is necessary to focus on the process of the moving entity in each process link.

[0015] The optimization of the recording strategy for the video of the nth process or some processes corresponding to the kth camera includes, but is not limited to, adjusting the boundary of the polygon electronic fence drawn on the high-precision electronic map, dynamically adjusting the video recording time window based on big data analysis, adjusting the deployment position of the positioning base station of the positioning system, and adjusting the position of the positioning device on the moving entity.

[0016] Step 5.1) When optimizing the recording strategy for the video of the nth process or some processes corresponding to the kth camera, it is advisable to first select to adjust the boundary of the polygon electronic fence drawn on the high-precision electronic map. Specifically,

[0017] Step 5.1.1) If there is a lack of the moving entity entering the process stage and / or there is a lack of the first half of the processing and production stage of the moving entity, according to the entering direction of the moving entity, find the "entering line segment" on the boundary of the polygon electronic fence on the high-precision electronic map, and expand the range of the polygon electronic fence in the opposite direction of the moving entity's entry.

[0018] Step 5.1.2) If there is a lack of the second half of the processing and production stage of the moving entity and / or there is a lack of the moving entity leaving the process stage, according to the leaving direction of the moving entity, find the "leaving line segment" on the boundary of the polygon electronic fence on the high-precision electronic map, and expand the range of the polygon electronic fence in the direction of the moving entity's leaving.

[0019] Step 5.1.3) If there are a large number of redundant parts unrelated to the video of the nth process or some processes before the mobile entity enters the process, find the "entry line segment" on the boundary of the polygon electronic fence on the high-precision electronic map according to the entry direction of the mobile entity, and shrink the range of the polygon electronic fence in the direction of the mobile entity's entry.

[0020] Step 5.1.4) If there are a large number of redundant parts unrelated to the video of the nth process or some processes after the mobile entity leaves the process, find the "exit line segment" on the boundary of the polygon electronic fence on the high-precision electronic map according to the exit direction of the mobile entity, and shrink the range of the polygon electronic fence in the opposite direction of the mobile entity's exit.

[0021] Step 5.2) Optimizing the recording strategy for the video of the nth process or some processes corresponding to the kth camera preferably dynamically adjusts the video recording time window based on big data analysis. Specifically,

[0022] Step 5.2.1) By tracing back a certain number of videos of the nth process or some processes, sampling big data analysis methods and non-linear function model modeling, estimate the actual process duration of the nth process or some processes.

[0023] Step 5.2.2) If there is a lack in the stage when the mobile entity enters the process and / or in the first half of the processing and production stage of the mobile entity, calculate the video duration T1 to be supplemented according to the estimated actual process duration of the nth process or some processes, and advance the original start time of video recording by T1. That is, the new start time of the optimized video recording is equal to the original start time minus T1.

[0024] Step 5.2.3) If there is a lack in the second half of the processing and production stage of the mobile entity and / or in the stage when the mobile entity leaves the process, calculate the video duration T2 to be supplemented according to the estimated actual process duration of the nth process or some processes, and delay the original end time of video recording by T2. That is, the new end time of the optimized video recording is equal to the original end time plus T2.

[0025] Step 5.2.4) If there are a large number of redundant parts unrelated to the video of the nth process or some processes before the mobile entity enters the process, calculate the video duration T3 to be deleted according to the estimated actual process duration of the nth process or some processes, and delay the original start time of video recording by T3. That is, the new start time of the optimized video recording is equal to the original start time plus T3.

[0026] Step 5.2.5) If there are a large number of redundant parts unrelated to the video of the nth process or certain processes after the mobile entity leaves the process for the video of the nth process or certain processes, calculate the video duration T4 to be deleted according to the estimated actual process duration of the nth process or certain processes, and advance by T4 at the original end time of video recording, that is, the new end time of the optimized video recording is equal to the original end time minus T4.

[0027] Step 5.3) The implementation of the recording strategy optimization for the video of the nth process or certain processes corresponding to the kth camera also includes adjusting the deployment position of the positioning base station of the positioning system and the way of adjusting the position of the positioning device on the mobile entity, so that the positioning device and the surrounding positioning base stations can avoid various obstacles as much as possible and maintain a line-of-sight distance.

[0028] Step 6: After determining through Step 5, if the video integrity of the nth process or certain processes corresponding to the kth camera is better than the video integrity threshold set by the system.

[0029] Step 7: The intelligent manufacturing factory management system platform stores the optimized process videos corresponding to all N processes experienced by any mobile entity and forms a complete process video sequence for any mobile entity, that is, the big data video in the whole life cycle during the production process, for inspection and traceability, so as to achieve the purpose of quality control and production efficiency improvement.

[0030] The present invention proposes a process video optimization method based on high-precision positioning, which combines video linkage with high-precision positioning technology to obtain the process video information of the corresponding production equipment by calculating the position information of the production equipment. However, in complex factory assembly line production operations, there are a large number of mechanical equipment, robotic arms, construction vehicles, workers, etc. The complex environment will block the positioning signal, as well as other random influences, causing the position information calculated by the system to jitter and drift, seriously affecting the reliability and accuracy of positioning, and further affecting the accuracy and completeness of the process video recording. In order to address key issues such as whether the collected video can truly reflect the actual situation of each process and whether the video is complete, the present invention introduces the indicator of video completeness to quantify the degree of integrity of the process video. When the video integrity is lower than the video integrity threshold set by the system, solutions such as adjusting the video area or dynamically adjusting the video recording time can be used to supplement the missing video information or delete the redundant video information, effectively improving the matching degree and integrity of the process video with the actual production situation, which can not only control the production quality of each process point on the production line and improve production efficiency, but also contribute to the big data analysis and mining of the process video of Industry 4.0 intelligent manufacturing, bringing huge direct and indirect economic value. The present invention will have broad application prospects for discrete manufacturing enterprises, as well as future fields such as Industry 4.0 intelligent manufacturing and smart factories. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A schematic diagram of the overall flow of the process video optimization method of the present invention.

[0032] Figure 2 Shows the relationship between a process video area on the production line and the entry and exit segments

[0033] Figure 3 This is a detailed flow chart of the optimization process video area (polygonal electronic fence boundary)

[0034] Figure 4 This is a specific flow chart for optimizing and dynamically adjusting the process video recording time window based on big data analysis

[0035] Figure 5 This is a schematic diagram of the process video area adjustment of Example 1

[0036] Figure 6 This is a schematic diagram of the process video area adjustment of Example 2

[0037] Figure 7 This is a schematic diagram of the process video area adjustment of Example 3

[0038] Figure 8 This is a schematic diagram of the process video area adjustment of Example 4 Specific Embodiments

[0039] Embodiment 1:

[0040] On a production line of a certain mechanical component, there are 120 production processes in total, and one of the core production processes is the 98th process; assume that there are 200 process videos of different mechanical components for tracing back the 98th process, among which 46 videos are incomplete. According to the definition of video integrity, the calculated integrity of the process video is 77%, which is lower than the system-set video integrity threshold of 90%; by analyzing the 46 incomplete videos, there is a 90% probability that there is a lack when a certain mechanical component enters the process stage. Therefore, it is selected to adjust the boundary of the polygon electronic fence drawn on the high-precision electronic map and expand the range of the polygon electronic fence in the opposite direction of the entry of the moving entity, as Figure 5 shown. Through the optimized adjustment of the video frame range, 41 incomplete videos are optimized into complete videos, and the integrity of the optimized process video is 97.5%. This embodiment only takes the 98th process as an example for illustration. For any other process or any combination of multiple processes, this method can still be used for optimization and is within the protection scope of the present invention.

[0041] Embodiment 2:

[0042] On a production line of a certain mechanical component, there are 180 production processes in total, and one of the core production processes is the 160th process; assume that there are 100 process videos of different mechanical components for tracing back the 160th process, among which 24 videos are incomplete. According to the definition of video integrity, the calculated integrity of the process video is 76%, which is lower than the system-set video integrity threshold of 85%; by analyzing the 24 incomplete videos, there is an 80% probability that there is a lack when a certain mechanical component leaves the process stage. Therefore, it is selected to adjust the boundary of the polygon electronic fence drawn on the high-precision electronic map and expand the range of the polygon electronic fence in the direction of the departure of a certain mechanical component, as Figure 6 shown. Through the optimized adjustment of the video frame range, 19 incomplete videos are optimized into complete videos, and the integrity of the optimized process video is 95%. This embodiment only takes the 160th process as an example for illustration. For any other process or any combination of multiple processes, this method can still be used for optimization and is within the protection scope of the present invention.

[0043] Embodiment 3:

[0044] On a production line of a certain mechanical component, there are 120 production processes in total, and one of the core production processes is the 115th process; assume that there are 200 process videos of different mechanical components for tracing back the 115th process, among which 33 videos have a large amount of redundancy. According to the definition of video integrity, the calculated integrity of the process video is 83.5%, which is lower than the system-set video integrity threshold of 90%; by analyzing the 33 abnormal process videos, there is an 85% probability that there is a large amount of video redundancy before a certain mechanical component enters the process. Therefore, it is selected to adjust the boundary of the polygon electronic fence drawn on the high-precision electronic map and narrow the range of the polygon electronic fence in the direction of the entry of a certain mechanical component, as Figure 7 shown. Through the optimized adjustment of the video frame range, 28 incomplete videos are optimized into complete videos, and the integrity of the optimized process video is 97.5%. This embodiment only takes the 115th process as an example for illustration. For any other process or any combination of multiple processes, this method can still be used for optimization and is within the protection scope of the present invention.

[0045] Example 4:

[0046] On a production line of a certain mechanical component, there are 180 production processes in total, and one of the core production processes is the 90th process; assume that there are 150 process videos of different mechanical components for tracing back the 90th process, among which 26 videos have a large amount of redundancy. According to the definition of video integrity, the calculated integrity of the process video is 82.7%, which is lower than the system-set video integrity threshold of 90%; by analyzing the 26 incomplete videos, there is a 90% probability that there is a large amount of video redundancy after a certain mechanical component leaves the process. Therefore, it is selected to adjust the boundary of the polygon electronic fence drawn on the high-precision electronic map and narrow the range of the polygon electronic fence in the opposite direction of the departure of the moving entity, as Figure 8 shown. Through the optimized adjustment of the video frame range, 23 incomplete videos are optimized into complete videos, and the integrity of the optimized process video is 98%. This embodiment only takes the 90th process as an example for illustration. For any other process or any combination of multiple processes, this method can still be used for optimization and is within the protection scope of the present invention.

[0047] Example 5:

[0048] On a production line of a certain mechanical component, there are a total of 120 production processes. One of the core production processes is the 98th process. Assume that there are 200 process videos of different mechanical components traced back to the 98th process, among which 46 videos are incomplete. According to the definition of video integrity, the calculated integrity of the process video is 77%, which is lower than the system-set video integrity threshold of 90%. By analyzing the 46 incomplete videos, there is a 90% probability that there is a lack when a certain mechanical component enters the process stage. In this embodiment, a method of dynamically adjusting the video recording time window based on big data analysis is selected. First, the big data analysis method and the non-linear function model are modeled to estimate that the actual average process duration of the 98th process is 28 seconds, while the current average process duration of video recording is only 20 seconds. Therefore, the video duration T1 to be supplemented is 8 seconds. Advance T1 (8 seconds) from the original start time of the 98th process video recording, that is, the new start time of the 98th process video recording after optimization is equal to the original start time minus 8 seconds. Through the processing of this method, 43 incomplete videos are optimized into complete videos, and the integrity of the process video after optimization is 98.5%. This embodiment only takes the 98th process as an example for illustration. For any other process, or any combination of multiple processes, this method can still be used for optimization, which is within the protection scope of the present invention.

[0049] Example 6:

[0050] On a production line of a certain mechanical component, there are a total of 180 production processes. One of the core production processes is the 160th process. Assume that there are 100 process videos of different mechanical components traced back to the 160th process, among which 24 videos are incomplete. According to the definition of video integrity, the calculated integrity of the process video is 76%, which is lower than the system-set video integrity threshold of 85%. By analyzing the 24 incomplete videos, there is an 80% probability that there is a lack when a certain mechanical component leaves the process stage. In this embodiment, a method of dynamically adjusting the video recording time window based on big data analysis is selected. First, the big data analysis method and the non-linear function model are modeled to estimate that the actual average process duration of the 160th process is 35 seconds, while the current average process duration of video recording is only 29 seconds. Therefore, the video duration T2 to be supplemented is 6 seconds. Delay T2 (6 seconds) from the original end time of the 160th process video recording, that is, the new end time of the video recording after optimization is equal to the original end time plus 6 seconds. Through the processing of this method, 20 incomplete videos are optimized into complete videos, and the integrity of the process video after optimization is 96%. This embodiment only takes the 160th process as an example for illustration. For any other process, or any combination of multiple processes, this method can still be used for optimization, which is within the protection scope of the present invention.

[0051] Example 7:

[0052] On a production line of a certain mechanical component, there are 120 production processes in total, and one of the core production processes is the 115th process; assume that there are 200 process videos of different mechanical components traced back to the 115th process, among which 33 videos have a large amount of redundancy. According to the definition of video integrity, the integrity of the process video is calculated to be 83.5%, which is lower than the system-set video integrity threshold of 90%; by analyzing the 33 abnormal process videos, there is an 85% probability that there is a large amount of video redundancy before a certain mechanical component enters the process. In this embodiment, a method of dynamically adjusting the video recording time window based on big data analysis is selected. First, the big data analysis method and the non-linear function model are modeled to estimate that the actual average process duration of the 115th process is 25 seconds, while the current average process duration of video recording is 38 seconds. Therefore, the video duration T3 to be deleted = 13 seconds. At the original start time of the 115th process video recording, lag by T3 (13 seconds), that is, the new start time of the optimized video recording is equal to the original start time plus 13 seconds. Through the processing of this method, 28 incomplete videos are optimized into complete videos, and the integrity of the optimized process video is 97.5%. This embodiment only takes the 115th process as an example for illustration. For any other process, or any combination of multiple processes, this method can still be used for optimization, which is within the protection scope of the present invention.

[0053] Embodiment 8:

[0054] On a production line of a certain mechanical component, there are 180 production processes in total, and one of the core production processes is the 90th process; assume that there are 150 process videos of different mechanical components traced back to the 90th process, among which 26 videos have a large amount of redundancy. According to the definition of video integrity, the integrity of the process video is calculated to be 82.7%, which is lower than the system-set video integrity threshold of 90%; by analyzing the 26 incomplete videos, there is a 90% probability that there is a large amount of video redundancy after a certain mechanical component leaves the process. In this embodiment, a method of dynamically adjusting the video recording time window based on big data analysis is selected. First, the big data analysis method and the non-linear function model are modeled to estimate that the actual average process duration of the 90th process is 45 seconds, while the current average process duration of video recording is 1 minute and 10 seconds. Therefore, the video duration T4 to be deleted = 25 seconds. At the original end time of the 90th process video recording, advance by T4 (25 seconds), that is, the new end time of the optimized video recording is equal to the original end time minus 25 seconds. Through the processing of this method, 23 incomplete videos are optimized into complete videos, and the integrity of the optimized process video is 98%. This embodiment only takes the 90th process as an example for illustration. For any other process, or any combination of multiple processes, this method can still be used for optimization, which is within the protection scope of the present invention.

[0055] In addition to the above embodiments, in the optimization of the process video, any step between step 5.1) and step 5.3) of the present invention, or a combination of any steps, is adopted to comprehensively optimize any process or the integrity of the process video of any combination of multiple processes, which is within the protection scope of the present invention.

Claims

1. A method for optimizing process videos based on high-precision positioning, characterized in that: Construct a high-precision electronic map of the production workshop in the smart factory, import the high-precision electronic map into the intelligent manufacturing factory management system platform, and draw a polygon electronic fence on the high-precision electronic map according to the actual physical space area of the nth process or several processes to represent the space area of the nth process or several processes; Bind the kth camera deployed on the production line to the number of the nth process or several processes, and at the same time map the recording area of the kth camera to the space area of the nth process or several processes; When any moving entity carrying a positioning device enters the space area of the nth process or several processes corresponding to the kth camera, the intelligent manufacturing factory management system platform will calculate the accurate position information of the moving entity carrying the positioning device and compare it with the polygon electronic fence representing the space area of the nth process or several processes on the high-precision electronic map; If the position information of the moving entity carrying the positioning device has entered the polygon electronic fence representing the space area of the nth process or several processes on the high-precision electronic map, the intelligent manufacturing factory management system platform will schedule the kth camera to start recording and synchronously store the corresponding process video until any of the moving entities leaves the space area of a certain process or several processes corresponding to the kth camera; Retrospect the videos of a certain number of moving entities entering the space area of the nth process (n = 1, 2,..., N) or several processes corresponding to the kth camera (k = 1, 2,..., K), and calculate the video integrity of the nth process or several processes corresponding to the kth camera; The video integrity of the nth process or several processes is expressed as where η n is the video integrity of the nth process or some processes, M is the total number of videos of the nth process or some processes for backtracking, and δ m,n is the video integrity identifier of the mth (m = 1, 2,..., M) video of the nth process or some processes; The process video includes three stages, namely the stage when the moving entity enters the process, the stage when the moving entity processes and produces, and the stage when the moving entity leaves the process; If the video of the nth process or several processes described in the mth item is completely matched with the actual process, then record δ m,n = 1; if the video of the nth process or several processes described in the mth item does not match the actual process, including but not limited to the lack of the moving entity entering the process stage, and / or the lack of the moving entity processing and producing stage, and / or the lack of the moving entity leaving the process stage, and / or there are a large number of irrelevant redundant parts in the video, then record δ m,n = 0; If the video integrity of the nth process or several processes corresponding to the kth camera is lower than the video integrity threshold set by the system, it is necessary to optimize the recording strategy for the video of the nth process or several processes corresponding to the kth camera, so that the video integrity of the nth process or several processes corresponding to the kth camera is better than the video integrity threshold set by the intelligent manufacturing factory management system platform; The optimization of the recording strategy for the video of the nth process or several processes corresponding to the kth camera includes, but is not limited to, adjusting the boundary of the polygon electronic fence drawn on the high-precision electronic map, dynamically adjusting the video recording time window based on big data analysis, adjusting the deployment position of the positioning base station of the positioning system, and adjusting the position of the positioning device on the moving entity; The intelligent manufacturing factory management system platform stores the optimized process videos corresponding to all N processes experienced by any moving entity and forms a complete process video sequence for any moving entity.

2. The method for optimizing process video based on high-precision positioning according to claim 1, wherein: The video integrity threshold is the threshold value of the process video integrity set for different moving entities, denoted as Γ, and its value range is between [0,1]; For moving entities with high technology maturity, the video integrity threshold Γ can be set to 60% - 70%; For moving entities with low technology maturity, the video integrity threshold Γ can be set to 85% - 95%.

3. The method for optimizing process video based on high-precision positioning according to claim 1, wherein: When optimizing the recording strategy for the video of the nth process or certain processes corresponding to the kth camera, it is preferably to first select to adjust the boundary of the polygon electronic fence drawn on the high-precision electronic map.

4. The method for optimizing process video based on high-precision positioning according to claim 3, wherein: The adjustment of the boundary of the polygon electronic fence drawn on the high-precision electronic map, specifically, If there are missing parts in the process stage when the moving entity enters, and / or there are missing parts in the first half of the processing and production stage of the moving entity, find the "entry line segment" on the boundary of the polygon electronic fence on the high-precision electronic map according to the entry direction of the moving entity, and expand the range of the polygon electronic fence in the opposite direction of the entry of the moving entity; If there are missing parts in the second half of the processing and production stage of the moving entity, and / or there are missing parts when the moving entity leaves the process stage, find the "exit line segment" on the boundary of the polygon electronic fence on the high-precision electronic map according to the exit direction of the moving entity, and expand the range of the polygon electronic fence in the direction of the exit of the moving entity; If there are a large number of redundant parts unrelated to the video of the nth process or certain processes in the video of the nth process or certain processes before the moving entity enters the process, find the "entry line segment" on the boundary of the polygon electronic fence on the high-precision electronic map according to the entry direction of the moving entity, and shrink the range of the polygon electronic fence in the direction of the entry of the moving entity; If there are a large number of redundant parts unrelated to the video of the nth process or certain processes in the video of the nth process or certain processes after the moving entity leaves the process, find the "exit line segment" on the boundary of the polygon electronic fence on the high-precision electronic map according to the exit direction of the moving entity, and shrink the range of the polygon electronic fence in the opposite direction of the exit of the moving entity.

5. The method for optimizing process video based on high-precision positioning according to claim 1, wherein: When optimizing the recording strategy for the video of the nth process or certain processes corresponding to the kth camera, it is preferably to secondarily select to dynamically adjust the video recording time window based on big data analysis.

6. The method for optimizing process video based on high-precision positioning according to claim 5, wherein: The dynamic adjustment of the video recording time window based on big data analysis, specifically, By backtracking the videos of a certain number of the nth process or some processes, sampling big data analysis methods and nonlinear function model modeling, estimate the actual process duration of the nth process or some processes; If there are missing parts in the process stage of the moving entity and / or in the first half of the processing and production stage of the moving entity, calculate the video duration T1 to be supplemented according to the estimated actual process duration of the nth process or some processes, and advance T1 from the original start time of video recording, that is, the new start time of the optimized video recording is equal to the original start time minus T1; If there are missing parts in the second half of the processing and production stage of the moving entity and / or in the process stage when the moving entity leaves the process, calculate the video duration T2 to be supplemented according to the estimated actual process duration of the nth process or some processes, and delay T2 from the original end time of video recording, that is, the new end time of the optimized video recording is equal to the original end time plus T2; If there are a large number of redundant parts unrelated to the video of the nth process or some processes before the moving entity enters the process in the video of the nth process or some processes, calculate the video duration T3 to be deleted according to the estimated actual process duration of the nth process or some processes, and delay T3 from the original start time of video recording, that is, the new start time of the optimized video recording is equal to the original start time plus T3; If there are a large number of redundant parts unrelated to the video of the nth process or some processes after the moving entity leaves the process in the video of the nth process or some processes, calculate the video duration T4 to be deleted according to the estimated actual process duration of the nth process or some processes, and advance T4 from the original end time of video recording, that is, the new end time of the optimized video recording is equal to the original end time minus T4.

7. A method for optimizing process videos based on high-precision positioning according to claim 1, characterized in that: The execution of recording strategy optimization for the video of the nth process or some processes corresponding to the kth camera also includes adjusting the deployment position of the positioning base station of the positioning system and the position mode of the positioning device on the moving entity, so that the positioning device and the surrounding positioning base stations avoid various obstacles as much as possible and maintain a line-of-sight distance.

Images