AI visual security inspection system

Abstract

The present invention relates to the technical field of regional security inspection, and discloses an AI visual security inspection system, comprising: based on analyzing real-time human flow data, obtaining key evaluation values, classifying and marking inspection sub-areas according to the key evaluation values, analyzing the marked key inspection sub-areas, obtaining key inspection sub-area groups and determining the initial setting number of inspection equipment, determining the cross-region inspection path of the inspection equipment according to the analysis of the key inspection sub-area groups, performing time period analysis on the key inspection sub-areas included in the key inspection sub-area groups, and determining the final number of inspection equipment and inspection methods. The present invention realizes rational inspection planning of inspection areas, and on this basis, effectively reduces the number of inspection equipment and realizes efficient inspection management.

Description

Technical Field

[0001] The present invention relates to the technical field of regional security inspection, and in particular to an AI visual security inspection system. Background Art

[0002] In recent years, the rapid development of artificial intelligence and computer vision technology has provided new solutions for security inspections. By combining AI vision technology, automated and intelligent security inspections can be achieved, inspection efficiency can be improved, inspection blind spots can be reduced, and rapid accident response can be achieved.

[0003] At present, in the process of realizing regional inspection through AI vision technology, the inspection equipment inspects all areas, which has the problem of low inspection efficiency and cannot timely monitor when abnormalities occur in the inspection area. In addition, in the current existing technology, regional inspections are also carried out through a large number of inspection equipment, resulting in high inspection resource costs. In addition, when no abnormalities occur in a large number of areas, the utilization rate of inspection resources is low. The inspection method is fixed, and the inspection resource costs are not effectively reduced and the inspection efficiency is not improved while realizing the inspection through the regulation of the inspection equipment and the optimization of the inspection method. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes an AI visual security inspection system.

[0005] The AI ​​visual security inspection system includes: a data acquisition module: obtaining real-time human flow data of the inspection sub-area every day in a historical period, wherein the real-time human flow data includes real-time human flow;

[0006] Area division module: Based on the analysis of real-time crowd flow data, the proportion of abnormal crowd flow duration is obtained. According to the proportion of abnormal crowd flow duration, the processing output is obtained to obtain the proportion of abnormal crowd flow days and the representation value of the abnormal crowd flow degree. The proportion of abnormal crowd flow duration and the representation value of the abnormal crowd flow degree are multiplied to obtain the key evaluation value. According to the key evaluation value, the inspection sub-area is classified and marked to obtain the key inspection sub-area and non-key inspection sub-area;

[0007] Inspection preliminary determination module: Based on the analysis of the marked key inspection sub-areas, the key inspection sub-area groups are obtained and the initial setting quantity of the inspection equipment is determined. Based on the analysis of the key inspection sub-area groups, the cross-area inspection paths of the inspection equipment are determined.

[0008] Inspection final determination module: conduct time period analysis on the key inspection sub-areas contained in the key inspection sub-area group to determine the final number of inspection equipment and inspection methods.

[0009] As a further technical solution of the present invention, the inspection sub-areas are classified and marked according to the key evaluation values ​​to obtain key inspection sub-areas and non-key inspection sub-areas, which specifically includes: comparing the key evaluation value with the key evaluation threshold; if the key evaluation value is greater than or equal to the key evaluation threshold, marking the inspection sub-area as a key inspection sub-area; if the key evaluation value is less than the key evaluation threshold, marking the inspection sub-area as a non-key inspection sub-area.

[0010] As a further technical solution of the present invention, the method for obtaining the proportion of abnormal duration of flow is as follows:

[0011] The real-time flow of people in the inspection sub-area every day in the historical period is compared with the preset real-time flow of people. If the real-time flow of people in the inspection sub-area every day in the historical period is greater than the preset real-time flow of people, a flow abnormality signal is generated;

[0012] The duration of abnormal crowd flow signals every day in the historical period is obtained, and it is compared with the duration of a day to obtain the proportion of abnormal crowd flow duration.

[0013] As a further technical solution of the present invention, the method of obtaining the abnormal duration ratio of abortion may further include:

[0014] The abnormal duration ratio of human flow is compared with the abnormal duration ratio threshold. If the abnormal duration ratio of human flow is greater than the abnormal duration ratio threshold, it means that the overall human flow on that day is abnormal. If the abnormal duration ratio of human flow is less than or equal to the abnormal duration ratio threshold, it means that the overall human flow on that day is normal.

[0015] During the historical period, the number of days with overall abnormal passenger flow is obtained, and the ratio of the number of days with overall abnormal passenger flow to the number of days in the historical period is processed to obtain the proportion of days with abnormal passenger flow.

[0016] As a further technical solution of the present invention, the method of obtaining the proportion of days with abnormal abortion may further include:

[0017] Perform difference processing on the abnormal duration ratio of the crowd flow corresponding to the overall abnormal crowd flow on that day and the abnormal duration ratio threshold of the crowd flow, and take the absolute value of the difference to obtain the abnormal duration ratio deviation value of the crowd flow;

[0018] The sum and average of all abnormal flow duration ratio deviation values ​​in the historical period are taken to obtain the mean deviation of abnormal flow duration ratio, and the mean deviation of abnormal flow duration ratio is processed with the abnormal flow duration ratio threshold to obtain the value representing the degree of abnormal flow.

[0019] As a further technical solution of the present invention, the key inspection sub-area groups are analyzed to determine the cross-area inspection path of the inspection equipment, specifically including:

[0020] Divide each key inspection sub-area and its adjacent inspection sub-area with the shortest distance into a key inspection sub-area group;

[0021] Outlining the two key inspection sub-areas included in the key inspection sub-area group to obtain boundary contour lines of the key inspection sub-areas, selecting a contour point on the boundary contour lines of the two key inspection sub-areas respectively, connecting them, and obtaining a cross-area inspection path;

[0022] The selection constraint of the contour points is: the distance between the contour points selected on the boundary contour lines of the two key inspection sub-areas is equal to the shortest distance between the boundary contour lines of the two key inspection sub-areas;

[0023] The initial number of inspection devices is equal to the number of key inspection sub-areas.

[0024] As a further technical solution of the present invention, the process of determining the final number of inspection equipment includes: obtaining the duration of the abnormal signal of the flow of people corresponding to the key inspection sub-area every day in the historical period, marking the time period corresponding to the duration as the abnormal time period, obtaining the overlap time period between the abnormal time periods corresponding to two key inspection sub-areas in the key inspection sub-area group, and marking the time length of the overlap time period as the abnormal overlap time;

[0025] The lengths of the abnormal time periods corresponding to the two key inspection sub-areas are summed to obtain the total abnormal time;

[0026] The abnormal coincidence time is processed with the ratio of the abnormal total time to obtain the abnormal coincidence time ratio;

[0027] Compare the abnormal coincidence time ratio with the abnormal coincidence time ratio threshold:

[0028] If the abnormal coincidence time ratio is greater than or equal to the abnormal coincidence time ratio threshold, no operation is performed;

[0029] If the abnormal coincidence time ratio is less than the abnormal coincidence time ratio threshold, a signal to be analyzed is generated.

[0030] As a further technical solution of the present invention, the process of determining the final number of inspection equipment also includes:

[0031] Based on the signal to be analyzed, the interval period between the abnormal periods corresponding to the two key inspection sub-areas is obtained, the time corresponding to the interval period is obtained, and it is marked as the interval time. The cross-area inspection path distance between the two key inspection sub-areas is obtained, and the cross-area inspection path distance is divided by the interval time to obtain the inspection speed, and the inspection speed is compared with the speed range of the inspection equipment:

[0032] If the inspection speed is within the speed range of the inspection equipment, the interval period is marked as a cross-zone inspection period;

[0033] If the inspection speed is not within the speed range of the inspection equipment, the interval period will be marked as a non-cross-zone inspection period.

[0034] As a further technical solution of the present invention, the process of determining the final number of inspection equipment also includes:

[0035] Count the number of time periods that can be inspected across different areas and the total number of intervals between abnormal time periods corresponding to two key inspection sub-areas, and perform ratio processing to obtain the ratio of the number of time periods that can be inspected across different areas;

[0036] Compare the ratio of the number of time periods that can be inspected across different zones with the threshold of the ratio of the number of time periods that can be inspected across different zones:

[0037] If the ratio of the number of time periods that can be inspected across different zones is greater than or equal to the ratio threshold of the number of time periods that can be inspected across different zones, the initial setting number of the inspection equipment is reduced by 1;

[0038] If the ratio of the number of time periods that can be inspected across zones is less than the ratio threshold of the number of time periods that can be inspected across zones, no operation is performed.

[0039] As a further technical solution of the present invention, the process of determining the inspection mode is as follows:

[0040] If the number of inspection devices corresponding to the key inspection sub-region group is 1, the key inspection sub-region group is marked as the first key inspection sub-region group; if the number of inspection devices corresponding to the key inspection sub-region group is 2, the key inspection sub-region group is marked as the second key inspection sub-region group;

[0041] For the first key inspection sub-region group, the inspection equipment corresponding to the first key inspection sub-region group only performs cross-region inspection on the two key inspection sub-regions included in the first key inspection sub-region group during the cross-region inspection period, wherein the inspection equipment needs to move according to the cross-region inspection path;

[0042] For the second key inspection sub-region group, the two key inspection sub-regions included in the second key inspection sub-region group can only be inspected by separate corresponding inspection equipment.

[0043] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0044] The real-time human flow data of the inspection sub-area is obtained every day in the historical period, and the key evaluation values ​​are obtained based on the analysis of the real-time human flow data. The inspection sub-areas are classified and marked according to the key evaluation values. The present invention divides the inspection sub-areas and judges the abnormal human flow in the inspection sub-areas, thereby determining the inspection scope, improving the security of the inspection area, and effectively avoiding the problem of low inspection efficiency in the inspection area. Based on the analysis of the marked key inspection sub-areas, the key inspection sub-area group is obtained and the initial setting number of inspection equipment is determined. According to the analysis of the key inspection sub-area group, the cross-region inspection path of the inspection equipment is determined, and the key inspection sub-areas included in the key inspection sub-area group are analyzed for time periods to determine the final number of inspection equipment and the inspection method. The present invention realizes the rationalization of inspection planning for the inspection area, and on this basis, effectively reduces the number of inspection equipment and realizes efficient inspection management. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1 is a flow chart of marking the division of inspection sub-areas provided by the first embodiment of the present invention;

[0047] Figure 2 is a flowchart of an AI visual security inspection system provided by a second embodiment of the present invention;

[0048] Figure 3 This is a flowchart for determining the number of inspection devices provided by the second embodiment of the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] In the description of the present invention, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., and orientations or positional relationships indicated are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0051] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.

[0052] In addition, in the description of the present specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0053] References to "one embodiment" or "some embodiments" etc. described in the present specification mean that one or more embodiments of the present invention include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0054] Embodiment 1

[0055] like Figure 1 As shown, the AI ​​visual security inspection system described in the embodiment of the present invention includes:

[0056] Data acquisition module: obtains real-time human flow data of the inspection sub-area every day in the historical period, wherein the real-time human flow data includes real-time human flow volume;

[0057] It should be noted that the pedestrian flow in the embodiment of the present invention is collected by AI vision technology. The advantage of using AI vision technology to collect pedestrian flow is that it can accurately identify and count pedestrians in the video, avoiding the errors that may exist in traditional manual counting. At the same time, it can also handle problems such as multi-person overlap and occlusion in complex scenes, and can be customized and expanded according to actual needs to meet application requirements in different scenarios.

[0058] The length of the historical cycle includes but is not limited to 10 days, 15 days or 20 days;

[0059] In some embodiments, the entire inspection area is divided into a number of sub-areas of equal area, and the sub-areas are marked as inspection sub-areas;

[0060] Area division module: Based on the analysis of real-time human flow data, key evaluation values ​​are obtained, and inspection sub-areas are classified and marked according to the key evaluation values, wherein the inspection sub-areas are classified and marked into key inspection sub-areas and non-key inspection sub-areas;

[0061] Exemplarily, the process of classifying and marking the inspection sub-areas is specifically as follows:

[0062] Obtain the real-time flow of people in the inspection sub-area every day in the historical period, and analyze it to obtain the proportion of abnormal duration of people flow. According to the proportion of abnormal duration of people flow, process and output the proportion of days with abnormal flow of people and the representation value of the degree of abnormal flow of people flow. Multiply the proportion of duration of abnormal flow of people flow and the representation value of abnormal degree of flow of people flow to obtain the key evaluation value.

[0063] Compare the key evaluation value with the key evaluation threshold. If the key evaluation value is greater than or equal to the key evaluation threshold, mark the inspection sub-area as a key inspection sub-area. If the key evaluation value is less than the key evaluation threshold, mark the inspection sub-area as a non-key inspection sub-area.

[0064] For example, the method for obtaining the proportion of abnormal flow duration is as follows:

[0065] The real-time flow of people in the inspection sub-area every day in the historical period is compared with the preset real-time flow of people. If the real-time flow of people in the inspection sub-area every day in the historical period is greater than the preset real-time flow of people, an abnormal flow signal is generated; if the real-time flow of people in the inspection sub-area every day in the historical period is less than or equal to the preset real-time flow of people, a normal flow signal is generated;

[0066] Obtain the duration of abnormal traffic flow signals every day in the historical period, and compare it with the duration of a day to obtain the proportion of abnormal traffic flow duration;

[0067] It should be noted that the preset real-time flow rate is set by those skilled in the art based on historical experimental data, and a day represents 24 hours;

[0068] For example, the method for obtaining the percentage of days with abnormal flow of people is as follows:

[0069] The abnormal duration ratio of human flow is compared with the abnormal duration ratio threshold. If the abnormal duration ratio of human flow is greater than the abnormal duration ratio threshold, it means that the overall human flow on that day is abnormal. If the abnormal duration ratio of human flow is less than or equal to the abnormal duration ratio threshold, it means that the overall human flow on that day is normal.

[0070] In the historical period, the number of days with overall abnormal passenger flow is obtained, and the ratio of the number of days with overall abnormal passenger flow to the number of days in the historical period is processed to obtain the proportion of days with abnormal passenger flow;

[0071] Exemplarily, the method for obtaining the value representing the degree of abnormality of human flow is as follows:

[0072] Perform difference processing on the abnormal duration ratio of the crowd flow corresponding to the overall abnormal crowd flow on that day and the abnormal duration ratio threshold, and take the absolute value of the difference to obtain the abnormal duration ratio deviation value of the crowd flow, sum and average all the abnormal duration ratio deviation values ​​of the crowd flow in the historical period to obtain the abnormal duration ratio deviation mean, perform ratio processing on the abnormal duration ratio deviation mean and the abnormal duration ratio threshold to obtain the abnormal degree representation value of the crowd flow;

[0073] The technical solution of the embodiment of the present invention is: obtaining real-time pedestrian flow data of the inspection sub-area every day within the historical period, obtaining key evaluation values ​​based on the analysis of the real-time pedestrian flow data, and classifying and marking the inspection sub-area according to the key evaluation values. The present invention divides the inspection sub-area, judges the abnormal pedestrian flow in the inspection sub-area, thereby determining its inspection range, improving the security of the inspection area, and effectively avoiding the problem of low inspection efficiency in the inspection area.

[0074] Embodiment 2

[0075] like Figure 2 As shown, based on Example 1, the AI ​​visual security inspection system described in this embodiment of the present invention includes:

[0076] Inspection preliminary determination module: Based on the analysis of the marked key inspection sub-areas, the key inspection sub-area groups are obtained and the initial setting quantity of the inspection equipment is determined. Based on the analysis of the key inspection sub-area groups, the cross-area inspection paths of the inspection equipment are determined.

[0077] It should be noted that the cross-region inspection path refers to the movement path of the inspection equipment from one key inspection sub-region to another key inspection sub-region;

[0078] In some embodiments, each key inspection sub-area and its adjacent inspection sub-area with the shortest distance are divided into a key inspection sub-area group;

[0079] Outlining the two key inspection sub-areas included in the key inspection sub-area group to obtain boundary contour lines of the key inspection sub-areas, selecting a contour point on the boundary contour lines of the two key inspection sub-areas respectively, connecting them, and obtaining a cross-area inspection path;

[0080] The selection constraint of the contour points is: the distance between the contour points selected on the boundary contour lines of the two key inspection sub-areas is equal to the shortest distance between the boundary contour lines of the two key inspection sub-areas;

[0081] Exemplarily, the process of obtaining the number of inspection devices includes:

[0082] Count the number of key inspection sub-areas, set the number of inspection equipment according to the number of key inspection sub-areas, and obtain the initial setting number of inspection equipment;

[0083] Inspection final determination module: conducts time period analysis on the key inspection sub-areas contained in the key inspection sub-area group to determine the final number of inspection equipment and inspection methods;

[0084] Specifically, Figure 3 As shown in the figure, the process of determining the final number of inspection equipment includes:

[0085] S1, obtain the duration of the abnormal signal of the flow of people corresponding to the key inspection sub-area every day in the historical period, mark the time period corresponding to the duration as the abnormal time period, obtain the overlap time period between the abnormal time periods corresponding to two key inspection sub-areas in the key inspection sub-area group, and mark the time length of the overlap time period as the abnormal overlap time;

[0086] The lengths of the abnormal time periods corresponding to the two key inspection sub-areas are summed to obtain the total abnormal time;

[0087] The abnormal coincidence time is processed with the ratio of the abnormal total time to obtain the abnormal coincidence time ratio;

[0088] Compare the abnormal coincidence time ratio with the abnormal coincidence time ratio threshold:

[0089] If the abnormal coincidence time ratio is greater than or equal to the abnormal coincidence time ratio threshold, no operation is performed;

[0090] If the abnormal coincidence time ratio is less than the abnormal coincidence time ratio threshold, a signal to be analyzed is generated;

[0091] S2, based on the signal to be analyzed, obtain the interval period between the abnormal periods corresponding to the two key inspection sub-areas respectively, obtain the time corresponding to the interval period, mark it as the interval time, obtain the cross-area inspection path distance between the two key inspection sub-areas, divide the cross-area inspection path distance by the interval time, obtain the inspection speed, and compare the inspection speed with the speed range of the inspection equipment:

[0092] If the inspection speed is within the speed range of the inspection equipment, the interval period is marked as a cross-zone inspection period;

[0093] If the inspection speed is not within the speed range of the inspection equipment, the interval period is marked as a non-cross-zone inspection period;

[0094] Count the number of time periods that can be inspected across different areas and the total number of intervals between abnormal time periods corresponding to two key inspection sub-areas, and perform ratio processing to obtain the ratio of the number of time periods that can be inspected across different areas;

[0095] Compare the ratio of the number of time periods that can be inspected across different zones with the threshold of the ratio of the number of time periods that can be inspected across different zones:

[0096] If the ratio of the number of time periods that can be inspected across different zones is greater than or equal to the ratio threshold of the number of time periods that can be inspected across different zones, the initial setting number of the inspection equipment is reduced by 1;

[0097] If the ratio of the number of time periods that can be inspected across different zones is less than the ratio threshold of the number of time periods that can be inspected across different zones, no operation will be performed;

[0098] It should be noted that if the ratio of the number of cross-regional inspection time periods is greater than or equal to the ratio threshold of the number of cross-regional inspection time periods, the idea of ​​reducing the number of inspection devices by 1 is that if the number of cross-regional inspection time periods is relatively large, it means that the inspection device can perform cross-regional inspections between key inspection sub-areas more often, and there is no need to use two inspection devices to perform corresponding separate inspections on the two key inspection sub-areas, which can improve the resource utilization rate of the inspection and reduce the inspection resource cost;

[0099] After analyzing all key inspection sub-area groups, the final number of inspection equipment is obtained;

[0100] The process of determining the inspection method is as follows:

[0101] If the number of inspection devices corresponding to the key inspection sub-region group is 1, the key inspection sub-region group is marked as the first key inspection sub-region group; if the number of inspection devices corresponding to the key inspection sub-region group is 2, the key inspection sub-region group is marked as the second key inspection sub-region group;

[0102] For the first key inspection sub-region group, the inspection equipment corresponding to the first key inspection sub-region group only performs cross-region inspection on the two key inspection sub-regions included in the first key inspection sub-region group during the cross-region inspection period, wherein the inspection equipment needs to move according to the cross-region inspection path;

[0103] For the second key inspection sub-area group, the two key inspection sub-areas contained in the second key inspection sub-area group can only be inspected by the corresponding inspection equipment alone;

[0104] It should be noted that, because the number of inspection devices initially set is the same as the number of key inspection sub-areas, and the number of inspection devices in the second key inspection sub-area group is not reduced, the key inspection sub-areas have separate corresponding inspection devices;

[0105] The technical solution of the embodiment of the present invention is: based on the analysis of the marked key inspection sub-areas, the key inspection sub-area group is obtained and the initial setting number of inspection equipment is determined; according to the analysis of the key inspection sub-area group, the cross-area inspection path of the inspection equipment is determined; the key inspection sub-areas included in the key inspection sub-area group are analyzed for a period of time to determine the final number of inspection equipment and the inspection method. The present invention realizes the rationalization of inspection planning for the inspection area, and on this basis, effectively reduces the number of inspection equipment and realizes efficient inspection management.

[0106] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. AI visual security inspection system, characterized by: include: Data acquisition module: obtains real-time human flow data of the inspection sub-area every day in the historical period, wherein the real-time human flow data includes real-time human flow volume; Area division module: Based on the analysis of real-time crowd flow data, the proportion of abnormal crowd flow duration is obtained. According to the proportion of abnormal crowd flow duration, the processing output is obtained to obtain the proportion of abnormal crowd flow days and the representation value of the abnormal crowd flow degree. The proportion of abnormal crowd flow duration and the representation value of the abnormal crowd flow degree are multiplied to obtain the key evaluation value. According to the key evaluation value, the inspection sub-area is classified and marked to obtain the key inspection sub-area and non-key inspection sub-area; Inspection preliminary determination module: Based on the analysis of the marked key inspection sub-areas, the key inspection sub-area groups are obtained and the initial setting quantity of the inspection equipment is determined. The key inspection sub-area groups are analyzed to determine the cross-area inspection paths of the inspection equipment. Inspection final determination module: conducts time period analysis on the key inspection sub-areas contained in the key inspection sub-area group to determine the final number of inspection equipment and inspection methods; The process of determining the final number of inspection equipment includes: Obtain the duration of the abnormal signal of the flow of people corresponding to the key inspection sub-area every day in the historical period, mark the period corresponding to the duration as the abnormal period, obtain the overlapping period between the abnormal periods corresponding to the two key inspection sub-areas in the key inspection sub-area group, and mark the length of the overlapping period as the abnormal overlapping time; The lengths of the abnormal time periods corresponding to the two key inspection sub-areas are summed to obtain the total abnormal time; The abnormal coincidence time is processed with the ratio of the abnormal total time to obtain the abnormal coincidence time ratio; Compare the abnormal coincidence time ratio with the abnormal coincidence time ratio threshold: If the abnormal coincidence time ratio is greater than or equal to the abnormal coincidence time ratio threshold, no operation is performed; If the abnormal coincidence time ratio is less than the abnormal coincidence time ratio threshold, a signal to be analyzed is generated; Based on the signal to be analyzed, the interval period between the abnormal periods corresponding to the two key inspection sub-areas is obtained, the time corresponding to the interval period is obtained, and it is marked as the interval time. The cross-area inspection path distance between the two key inspection sub-areas is obtained, and the cross-area inspection path distance is divided by the interval time to obtain the inspection speed, and the inspection speed is compared with the speed range of the inspection equipment: If the inspection speed is within the speed range of the inspection equipment, the interval period is marked as a cross-zone inspection period; If the inspection speed is not within the speed range of the inspection equipment, the interval period is marked as a non-cross-zone inspection period; Count the number of time periods that can be inspected across different areas and the total number of intervals between abnormal time periods corresponding to two key inspection sub-areas, and perform ratio processing to obtain the ratio of the number of time periods that can be inspected across different areas; Compare the ratio of the number of time periods that can be inspected across different zones with the threshold of the ratio of the number of time periods that can be inspected across different zones: If the ratio of the number of time periods that can be inspected across different zones is greater than or equal to the ratio threshold of the number of time periods that can be inspected across different zones, the initial setting number of the inspection equipment is reduced by 1; If the ratio of the number of time periods that can be inspected across zones is less than the ratio threshold of the number of time periods that can be inspected across zones, no operation is performed.

2. The AI ​​visual security inspection system according to claim 1 is characterized in that: The classification and marking of the inspection sub-areas according to the key evaluation values ​​to obtain the key inspection sub-areas and non-key inspection sub-areas specifically includes: The key evaluation value is compared with the key evaluation threshold. If the key evaluation value is greater than or equal to the key evaluation threshold, the inspection sub-area is marked as a key inspection sub-area. If the key evaluation value is less than the key evaluation threshold, the inspection sub-area is marked as a non-key inspection sub-area.

3. The AI ​​visual security inspection system according to claim 1 is characterized in that: The method for obtaining the proportion of abnormal flow duration is as follows: The real-time flow of people in the inspection sub-area every day in the historical period is compared with the preset real-time flow of people. If the real-time flow of people in the inspection sub-area every day in the historical period is greater than the preset real-time flow of people, a flow abnormality signal is generated; The duration of abnormal crowd flow signals every day in the historical period is obtained, and it is compared with the duration of a day to obtain the proportion of abnormal crowd flow duration.

4. The AI ​​visual security inspection system according to claim 3 is characterized in that: The abnormal duration ratio of abortion is obtained, and then the following is included: The abnormal duration ratio of human flow is compared with the abnormal duration ratio threshold. If the abnormal duration ratio of human flow is greater than the abnormal duration ratio threshold, it means that the overall human flow on that day is abnormal. If the abnormal duration ratio of human flow is less than or equal to the abnormal duration ratio threshold, it means that the overall human flow on that day is normal. During the historical period, the number of days with overall abnormal passenger flow is obtained, and the ratio of the number of days with overall abnormal passenger flow to the number of days in the historical period is processed to obtain the proportion of days with abnormal passenger flow.

5. The AI ​​visual security inspection system according to claim 4 is characterized in that: The above-mentioned percentage of days with abnormal abortion also includes: Perform difference processing on the abnormal duration ratio of the crowd flow corresponding to the overall abnormal crowd flow on that day and the abnormal duration ratio threshold of the crowd flow, and take the absolute value of the difference to obtain the abnormal duration ratio deviation value of the crowd flow; The sum and average of all abnormal flow duration ratio deviation values ​​in the historical period are taken to obtain the mean deviation of abnormal flow duration ratio, and the mean deviation of abnormal flow duration ratio is processed with the abnormal flow duration ratio threshold to obtain the value representing the degree of abnormal flow.

6. The AI ​​visual security inspection system according to claim 1, characterized in that: The analysis of the key inspection sub-region groups to determine the cross-region inspection paths of the inspection equipment specifically includes: Divide each key inspection sub-area and its adjacent inspection sub-area with the shortest distance into a key inspection sub-area group; Outlining the two key inspection sub-areas included in the key inspection sub-area group to obtain boundary contour lines of the key inspection sub-areas, selecting a contour point on the boundary contour lines of the two key inspection sub-areas respectively, connecting them, and obtaining a cross-area inspection path; The selection constraint of the contour points is: the distance between the contour points selected on the boundary contour lines of the two key inspection sub-areas is equal to the shortest distance between the boundary contour lines of the two key inspection sub-areas; The initial number of inspection devices is equal to the number of key inspection sub-areas.

7. The AI ​​visual security inspection system according to claim 1, characterized in that: The process of determining the inspection method is as follows: If the number of inspection devices corresponding to the key inspection sub-region group is 1, the key inspection sub-region group is marked as the first key inspection sub-region group; if the number of inspection devices corresponding to the key inspection sub-region group is 2, the key inspection sub-region group is marked as the second key inspection sub-region group; For the first key inspection sub-region group, the inspection equipment corresponding to the first key inspection sub-region group only performs cross-region inspection on the two key inspection sub-regions included in the first key inspection sub-region group during the cross-region inspection period, wherein the inspection equipment needs to move according to the cross-region inspection path; For the second key inspection sub-region group, the two key inspection sub-regions included in the second key inspection sub-region group can only be inspected by separate corresponding inspection equipment.

Images