Anti-trailing gate system based on image recognition

Abstract

The invention discloses an anti-trailing gate system based on image recognition, and relates to the technical field of intelligent anti-trailing gates. The anti-trailing gate system based on image recognition comprises a data acquisition module, a data processing module, a data processing module and a data processing module, wherein the data acquisition module is used for acquiring traffic area image data of a set traffic area in real time; the system comprises an identification analysis module, a judgment analysis module, an index analysis module, a comprehensive analysis module, an anti-following judgment module and an anti-following execution module which are used for acquiring image data of a passing area in real time and judging whether a passing person area exists in the set passing area or not according to an identification analysis result. The system can efficiently extract a passing person area, through combination of connected domain analysis and centroid coordinate calculation, the system can accurately position the center coordinates of passing persons, and then whether adjacent passing persons with abnormal spacing exist or not is judged through centroid spacing indexes.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent anti-tailgating gate machine, and in particular to an anti-tailgating gate machine system based on image recognition. Background Art

[0002] With the continuous improvement of social security management needs, anti-tailgating gate systems have been widely used in places such as subway stations, office buildings, and financial institutions that require strict access control. Traditional gate systems mainly use a single identity verification method (such as card swiping, fingerprint, face recognition, etc.) to determine whether the passerby has access rights. However, this technology has certain limitations and cannot effectively identify the phenomenon of tailgating. The tailgators often take advantage of the gaps that are not closed in time when the person in front passes through the gate to achieve unauthorized passage, which poses a hidden danger to the safety of the venue.

[0003] The current anti-tailgating gate system usually relies on simple identification of passers-by and static image comparison, and can only achieve basic pass and block functions. This method mainly relies on a single parameter judgment (such as whether there are passers-by) to detect tailgating behavior, and fails to conduct in-depth analysis of the dynamic state of the passage area. Therefore, in complex traffic scenarios, such as crowded traffic during peak hours and multiple people walking together, the system is prone to the following problems: response lag, misjudgment and missed judgment, and fixed threshold limitations.

[0004] Prior art, such as the invention patent application with announcement number: CN111723770B, discloses an anti-tailgating gate system and method based on image recognition, including a microprocessor, wherein the microprocessor is respectively connected with an image recognition module, an anti-tailgating peripheral, a special passage module and a power module, wherein the image recognition module recognizes the identity of the passer-by through an image and determines whether the passer-by can pass, and sends a signal of the recognition result to the microprocessor; the anti-tailgating peripheral, if the received result is pass, starts the anti-tailgating peripheral to open the gate; the special passage module recognizes the materials and companions of the passer-by through an image, determines whether the passer-by belongs to a preset special passage category, and sends the result to the microprocessor to block the signal of the image recognition module to start the anti-tailgating peripheral if the passer-by belongs to the special passage category, and does not send a signal to the microprocessor if the passer-by does not belong to the special passage category; by identifying the special case, the gate is controlled to release the special case to avoid special cases such as large luggage and children being blocked by mistake.

[0005] Based on the above scheme, it is found that the limitations of the existing technology include at least the following problems. First, the existing technology mainly relies on simple identity recognition and image comparison judgment of passers-by, and can only realize basic passing and blocking functions. It fails to conduct in-depth analysis of the dynamic state of the passage area. Therefore, it is difficult to capture the movement state and relative behavior of passers-by in real time, which easily leads to a delay in the system's response to the tailing behavior. Secondly, the basis for judging the tailing behavior in the existing technology is too single. In complex traffic scenarios (such as crowded traffic, fast passing, etc.), it is difficult to accurately distinguish between normal queuing and tailing phenomena, and the tailing behind bypasses the simple image recognition module by accelerating, approaching, etc., and it is difficult for the system to identify abnormal behavior in time. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present invention provides an anti-tailgating gate system based on image recognition, which solves the problems that the prior art does not deeply analyze the dynamic state of the passage area, making it difficult to capture the movement behavior of people in real time, resulting in a delayed response to the tailgating behavior, and the judgment basis is single, making it difficult to distinguish between normal passage and tailgating.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: an anti-tailgating gate system based on image recognition, comprising: a data acquisition module, an identification and analysis module, a judgment and analysis module, an index analysis module, a comprehensive analysis module, an anti-tailgating judgment module, and an anti-tailgating execution module; the data acquisition module is used to acquire the passage area image data of the set passage area in real time, and the passage area image data includes a number of pixels in the passage area image, and the pixel value and two-dimensional coordinates of each pixel; the identification and analysis module is used to identify and analyze the passage area image data of the set passage area in real time, and judge whether there is a passage area for passers-by in the set passage area according to the identification and analysis results; the judgment and analysis module is used to judge in real time whether there are adjacent passage areas with abnormal spacing in the set passage area when there is a passage area for passers-by in the set passage area. ; The index analysis module is used to analyze the centroid spacing index, speed difference index, path angle deviation index and contour area index of adjacent travelers with abnormal spacing in the set passage area when there are adjacent travelers with abnormal spacing in the set passage area; the comprehensive analysis module is used to comprehensively analyze the centroid spacing index, speed difference index, path angle deviation index and contour area index of adjacent travelers with abnormal spacing in the set passage area to obtain the tailing abnormality comprehensive index of adjacent travelers with abnormal spacing in the set passage area; the anti-tailgating judgment module is used to judge whether the tailing abnormality comprehensive index of adjacent travelers with abnormal spacing is higher than the tailing abnormality comprehensive index threshold; the anti-tailgating execution module is used to input preset anti-tailgating measures to the anti-tailgating gate when the tailing abnormality comprehensive index of adjacent travelers with abnormal spacing is higher than the tailing abnormality comprehensive index threshold.

[0008] Furthermore, the specific steps for judging whether there is a pedestrian area in the set passage area based on the recognition and analysis results are as follows: grayscale processing is performed on the passage area image data of the set passage area to obtain the passage area grayscale image data of the set passage area, wherein the passage area grayscale image data includes a number of grayscale pixels in the passage area grayscale image, and the grayscale pixel value of each grayscale pixel; based on a preset grayscale pixel interval, binarization processing is performed on the passage area grayscale image data of the set passage area to obtain a number of pedestrian grayscale pixels in the passage area grayscale image; if the number of pedestrian grayscale pixels in the passage area grayscale image is lower than the preset grayscale pixel number threshold, it indicates that there is no pedestrian area in the set passage area; if the number of pedestrian grayscale pixels in the passage area grayscale image is higher than or equal to the preset grayscale pixel number threshold, it indicates that there is a pedestrian area in the set passage area.

[0009] Furthermore, the specific steps for determining in real time whether there are adjacent passers-by with abnormal spacing in the set passage area are as follows: a connected domain analysis is performed on a number of grayscale pixel points of passers-by in the grayscale image of the passage area to obtain a number of grayscale regions of passers-by in the grayscale image of the passage area, and each grayscale region of passers-by includes a number of grayscale pixel points of passers-by; for each grayscale region of passers-by in the grayscale image of the passage area, the two-dimensional coordinates of each grayscale pixel point of passers-by in the grayscale region of passers-by are read respectively and a comprehensive analysis is performed to obtain the centroid coordinates of each grayscale region of passers-by in the grayscale image of the passage area; each grayscale region of passers-by in the grayscale image of the passage area is respectively compared with the two-dimensional coordinates of each grayscale pixel point of passers-by in the grayscale image of the passage area. The adjacent centroid spacing analysis is performed on each of the remaining grayscale areas of the passers-by to obtain the centroid spacing indexes of several groups of adjacent passers-by in the set passage area; it is judged whether the centroid spacing index of each group of adjacent passers-by in the set passage area is respectively within the preset centroid spacing threshold interval; if each centroid spacing index in the set passage area is within the preset centroid spacing threshold interval, it indicates that there are no adjacent passers-by with abnormal spacing in the set passage area; if there are adjacent passers-by in the set passage area whose centroid spacing index is outside the preset centroid spacing threshold interval, they are marked as adjacent passers-by with abnormal spacing; wherein, the specific formula for calculating the centroid coordinates of each grayscale area of ​​the passers-by in the grayscale image of the passage area is as follows: Among them, (x′ i , y′ i ) is the centroid coordinate of the grayscale area of ​​the i-th passerby in the grayscale image of the pass area, (x ir ,y ir) is the two-dimensional coordinate of the rth grayscale pixel point of the i-th grayscale area of ​​the passerby in the grayscale image of the passerby area, i=1, 2, 3, ..., i0, i0 is the number of grayscale areas of the passerby in the grayscale image of the passerby area.

[0010] Furthermore, the specific formula for calculating the centroid distance index of each group of adjacent passers-by in the set passage area is as follows: Among them, ZxJ t is the centroid distance index of the tth group of adjacent passers-by in the set passage area, (x′ t1 , y′ t1 ) is the centroid coordinate of the grayscale area of ​​the first passerby in the tth group of adjacent passersby in the grayscale image of the pass area, (x′ t2 , y′ t2 ) is the centroid coordinate of the grayscale area of ​​the second passerby in the tth group of adjacent passersby in the grayscale image of the passage area, t = 1, 2, 3, ..., t0, t0 is the number of adjacent passerby groups in the grayscale image of the passage area.

[0011] Furthermore, the specific formula for calculating the comprehensive index of tailing anomaly of adjacent passers-by with abnormal spacing in a set passage area is as follows: Among them, WsY is the comprehensive index of tailing abnormality of adjacent travelers with abnormal spacing in the set passage area, ZxJ is the center of mass spacing index of adjacent travelers with abnormal spacing in the set passage area, μ1 is the spacing coefficient stored in the database, ScY is the speed difference index of adjacent travelers with abnormal spacing in the set passage area, μ2 is the speed coefficient stored in the database, LjP is the path angle deviation index of adjacent travelers with abnormal spacing in the set passage area, μ3 is the angle deviation coefficient stored in the database, LmZ is the contour area index of adjacent travelers with abnormal spacing in the set passage area, μ4 is the contour area coefficient stored in the database, μ1+μ2+μ3+μ4=1, and π is pi.

[0012] Furthermore, when there are adjacent travelers with abnormal spacing in the set passage area, the specific steps of analyzing the speed difference index and path angle deviation index of the adjacent travelers with abnormal spacing in the set passage area are as follows: for two travelers among the adjacent travelers with abnormal spacing in the grayscale image of the passage area, the center of mass coordinates of several time points are continuously analyzed respectively, and a comprehensive analysis is performed to obtain the travel speed values ​​and speed vectors of the two travelers among the adjacent travelers with abnormal spacing in the grayscale image of the passage area; and the travel speed values ​​and speed vectors of the two travelers among the adjacent travelers with abnormal spacing in the grayscale image of the passage area are comprehensively analyzed to obtain the speed difference index and path angle deviation index of the adjacent travelers with abnormal spacing in the set passage area.

[0013] Furthermore, the specific formula for calculating the speed difference index and path angle deviation index of adjacent pedestrians with abnormal spacing in the set passage area is as follows: Among them, ScY is the speed difference index of adjacent travelers with abnormal spacing in the set passage area, e is a natural explanation, SdZ1 is the travel speed value of the first traveler among the adjacent travelers with abnormal spacing in the grayscale image of the passage area, SdZ2 is the travel speed value of the second traveler among the adjacent travelers with abnormal spacing in the grayscale image of the passage area, LjP is the path angle deviation index of adjacent travelers with abnormal spacing in the set passage area, is the velocity vector of the first passerby among the adjacent passersby with abnormal spacing in the grayscale image of the passing area, It is the velocity vector of the second passerby among the adjacent passersby with abnormal spacing in the grayscale image of the passage area.

[0014] Furthermore, when there are adjacent passers-by with abnormal spacing in the set passage area, the specific steps of analyzing the contour area index of the adjacent passers-by with abnormal spacing in the set passage area are as follows: the number of grayscale pixel points of the passers-by in the first and second grayscale areas of the adjacent passers-by with abnormal spacing in the grayscale image of the passage area are counted respectively, and they are marked as the contour area values ​​of the first and second grayscale areas of the adjacent passers-by with abnormal spacing in the grayscale image of the passage area, respectively, and a comprehensive analysis is performed to obtain the contour area index of the adjacent passers-by with abnormal spacing in the grayscale image of the passage area, and the calculation formula is as follows: Among them, LmZ is the contour area index of adjacent passers-by with abnormal spacing in the set passage area, LkM1 is the contour area value of the grayscale area of ​​the first passer-by among the adjacent passers-by with abnormal spacing in the grayscale image of the passage area, and LkM2 is the contour area value of the grayscale area of ​​the second passer-by among the adjacent passers-by with abnormal spacing in the grayscale image of the passage area.

[0015] Furthermore, the specific steps for obtaining the threshold value of the comprehensive index of tailgating anomaly are as follows: real-time statistics of the time interval values ​​of several people passing through the anti-tailgating gate, and comprehensive analysis are performed to obtain the average time interval; the time interval parameter value and the initial threshold value of the comprehensive index of tailgating anomaly are obtained, and a comprehensive analysis is performed in combination with the average time interval to obtain the threshold value of the comprehensive index of tailgating anomaly.

[0016] Furthermore, the specific formulas for calculating the mean of the travel time interval and the threshold of the comprehensive index of tailing anomaly are as follows: Among them, TxJ is the mean time interval of the passage, XjG sis the time interval value of the sth person passing through the anti-tailgating gate, ZhY is the tailgating abnormality comprehensive index threshold, ChY is the initial tailing abnormality comprehensive index threshold, α is the adjustment coefficient stored in the database, TxC is the travel time interval parameter value, s=1, 2, 3, …, s0, s0 is the number of people passing through the anti-tailgating gate.

[0017] The present invention has the following beneficial effects:

[0018] (1) The anti-tailgating gate system based on image recognition can efficiently extract the area of ​​​​passing personnel by acquiring image data of the passage area in real time and performing grayscale processing and binarization processing. Combined with the connected domain analysis and centroid coordinate calculation, the system can accurately locate the center coordinates of the passing personnel, and then determine whether there are adjacent passing personnel with abnormal spacing through the centroid spacing index. Compared with the simple static image judgment of the existing technology, this method can dynamically monitor the passage status, effectively identify tailgating behavior in complex scenarios, reduce the misjudgment and missed judgment rates, and improve the accuracy and reliability of tailgating detection.

[0019] (2) The anti-tailgating gate system based on image recognition calculates the centroid distance index, speed difference index, path angle deviation index and contour area index of adjacent travelers with abnormal spacing, and comprehensively analyzes the spatial position, movement speed, direction of travel and regional contour characteristics between people. The quantitative analysis of these parameters ensures that the system can fully capture the multi-dimensional characteristics of tailgating behavior. Through the comprehensive evaluation of the comprehensive index of tailing anomalies, the system can more accurately distinguish between tailgating phenomena and normal traffic behaviors, thus solving the problem of misjudgment of a single indicator.

[0020] (3) The anti-tailgating gate system based on image recognition dynamically optimizes the tailgating detection standard by real-time counting of the average time interval of the anti-tailgating gate and combining the preset time interval parameter value and the initial threshold. This method effectively adapts to different traffic scenarios, such as high traffic and low traffic, and improves the robustness and flexibility of the system. The dynamic threshold adjustment not only reduces the misjudgment of normal passers-by, but also enhances the practicality and traffic efficiency of the system in actual applications.

[0021] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a block diagram of an anti-tailgating gate system based on image recognition in the present invention.

[0023] Figure 2 This is a flowchart of the specific steps of determining whether there is a pedestrian area within a set passage area according to the recognition and analysis results in an anti-tailgating gate system based on image recognition of the present invention.

[0024] Figure 3 The present invention is a flowchart of the specific steps for obtaining a tailgating anomaly comprehensive index threshold in an anti-tailgating gate system based on image recognition. DETAILED DESCRIPTION

[0025] The overall idea of ​​the problem in the embodiment of this application is as follows:

[0026] Firstly, the image data of the set passage area is acquired in real time through the data acquisition module, including pixel points, grayscale values ​​and two-dimensional coordinates, and the image data are grayed and binarized to extract the grayscale area of ​​personnel in the passage area to determine whether there is a passage area for personnel. After confirming that there is a passage area for personnel, the connected domain analysis is performed on the passage personnel, the centroid coordinates of each passage personnel are calculated, and the centroid distance analysis is performed based on the centroid coordinates to identify adjacent passage personnel with abnormal spacing. For adjacent passage personnel with abnormal spacing, the centroid distance index, speed difference index, path angle deviation index and contour area index are further calculated, and the tailing abnormality comprehensive index is calculated through the comprehensive analysis module. Finally, the anti-tailing judgment module is used to determine whether the tailing abnormality comprehensive index exceeds the set threshold. If it exceeds the threshold, the system sends a signal to the anti-tailing execution module to trigger the anti-tailing measures to prevent unauthorized personnel from tailing and passing. In addition, the threshold of the tailing abnormality comprehensive index is dynamically adjusted by real-time statistics of the mean of the passage time interval, so that the system can adapt to different passage scenarios.

[0027] See also Figure 1The embodiment of the present invention provides a technical solution: an anti-tailgating gate system based on image recognition, comprising: a data acquisition module, an identification and analysis module, a judgment and analysis module, an index analysis module, a comprehensive analysis module, an anti-tailgating judgment module, and an anti-tailgating execution module; the data acquisition module is used to acquire the passage area image data of the set passage area in real time through a camera, and the passage area image data includes a number of pixels in the passage area image, and the pixel value and two-dimensional coordinates of each pixel; the identification and analysis module is used to identify and analyze the passage area image data of the set passage area in real time, and judge whether there is a passing personnel area in the set passage area according to the identification and analysis results; the judgment and analysis module is used to judge whether there are adjacent passing personnel with abnormal spacing in the set passage area in real time when there is a passing personnel area in the set passage area ... A numerical analysis module is used to analyze the centroid spacing index, speed difference index, path angle deviation index and contour area index of adjacent travelers with abnormal spacing in the set passage area when there are adjacent travelers with abnormal spacing in the set passage area; a comprehensive analysis module is used to comprehensively analyze the centroid spacing index, speed difference index, path angle deviation index and contour area index of adjacent travelers with abnormal spacing in the set passage area to obtain the tailing abnormal comprehensive index of adjacent travelers with abnormal spacing in the set passage area; an anti-tailgating judgment module is used to judge whether the tailing abnormal comprehensive index of adjacent travelers with abnormal spacing is higher than the tailing abnormal comprehensive index threshold; an anti-tailgating execution module is used to input preset anti-tailgating measures to the anti-tailgating gate when the tailing abnormal comprehensive index of adjacent travelers with abnormal spacing is higher than the tailing abnormal comprehensive index threshold.

[0028] Specifically, Figure 2 As shown, the specific steps of judging whether there is a pedestrian area in the set passage area according to the recognition and analysis results are as follows: grayscale processing is performed on the passage area image data of the set passage area to obtain the passage area grayscale image data of the set passage area, and the passage area grayscale image data includes a number of grayscale pixels in the passage area grayscale image, and the grayscale pixel value of each grayscale pixel; based on a preset grayscale pixel interval, binarization processing is performed on the passage area grayscale image data of the set passage area to obtain a number of pedestrian grayscale pixels in the passage area grayscale image, that is, the grayscale pixels whose grayscale pixel values ​​are in the preset grayscale pixel interval are marked as pedestrian grayscale pixels; if the number of pedestrian grayscale pixels in the passage area grayscale image is lower than the preset grayscale pixel number threshold, it indicates that there is no pedestrian area in the set passage area; if the number of pedestrian grayscale pixels in the passage area grayscale image is higher than or equal to the preset grayscale pixel number threshold, it indicates that there is a pedestrian area in the set passage area.

[0029] In this implementation scheme, by graying and binarizing the image data of the passage area, irrelevant background information can be effectively filtered out, the grayscale pixels of the passers-by can be accurately extracted, and the interference of factors such as ambient light and background clutter on the detection results can be reduced. By counting the number of grayscale pixels and comparing them with the preset grayscale pixel number threshold, it is possible to quickly and in real time determine whether there is a passing area in the set passage area. This method has clear logic and small calculation amount, is suitable for real-time systems, and improves detection efficiency. The combined judgment mechanism of grayscale pixel interval and pixel number threshold can effectively cope with complex passage scenes (such as light changes, partial occlusion, etc.), ensure that the presence of people in the passage area can be accurately judged under different conditions, and enhance the robustness of the system. On the basis of judging the existence of passers-by, the grayscale pixel area of ​​the passers-by is output, which provides accurate data input for subsequent multi-dimensional parameter analysis such as centroid coordinate calculation, spacing analysis, and speed detection, thereby ensuring the overall logical coherence and reliability of the system.

[0030] Specifically, the specific steps for determining in real time whether there are adjacent passers-by with abnormal spacing in the set passage area are as follows: a connected domain analysis is performed on a number of grayscale pixel points of passers-by in the grayscale image of the passage area to obtain a number of grayscale regions of passers-by in the grayscale image of the passage area, and each grayscale region of passers-by includes a number of grayscale pixel points of passers-by; for each grayscale region of passers-by in the grayscale image of the passage area, the two-dimensional coordinates of each grayscale pixel point of passers-by in the grayscale region of passers-by are read respectively and a comprehensive analysis is performed to obtain the centroid coordinates of each grayscale region of passers-by in the grayscale image of the passage area; each passers-by in the grayscale image of the passage area is connected to a plurality of grayscale regions of passers-by. The adjacent centroid distance analysis is performed on the grayscale area of ​​each passer-by respectively with the remaining grayscale area of ​​each passer-by in the grayscale image of the passage area to obtain the centroid distance index of several groups of adjacent passers-by in the set passage area; it is judged whether the centroid distance index of each group of adjacent passers-by in the set passage area is respectively within the preset centroid distance threshold interval; if each centroid distance index of the set passage area is within the preset centroid distance threshold interval, it indicates that there are no adjacent passers-by with abnormal distance in the set passage area; if there are adjacent passers-by in the set passage area whose centroid distance index is outside the preset centroid distance threshold interval, they are marked as adjacent passers-by with abnormal distance.

[0031] The specific formula for calculating the centroid coordinates of each grayscale area of ​​a passer-by in the grayscale image of the pass area is as follows: Among them, (x′ i , y′ i ) is the centroid coordinate of the grayscale area of ​​the i-th passerby in the grayscale image of the pass area, (x ir ,y ir) is the two-dimensional coordinate of the rth grayscale pixel point of the i-th grayscale area of ​​the passerby in the grayscale image of the passerby area, i = 1, 2, 3, ..., i0, i0 is the number of grayscale areas of the passerby in the grayscale image of the passerby area.

[0032] The specific formula for calculating the centroid distance index of each group of adjacent passers-by in the set passage area is as follows: Among them, ZxJ t is the centroid distance index of the tth group of adjacent passers-by in the set passage area, (x′ t1 , y′ t1 ) is the centroid coordinate of the grayscale area of ​​the first passerby in the tth group of adjacent passersby in the grayscale image of the pass area, (x′ t2 , y′ t2 ) is the centroid coordinate of the grayscale area of ​​the second passerby in the tth group of adjacent passersby in the grayscale image of the passage area, t = 1, 2, 3, ..., t0, t0 is the number of adjacent passerby groups in the grayscale image of the passage area.

[0033] In this implementation scheme, the grayscale areas of passers-by are extracted through connected domain analysis, and the centroid coordinates of each area are calculated, so as to accurately locate the position center of each passer-by. This method avoids recognition errors caused by image noise or edge interference, and ensures the accuracy of the position of passers-by. The actual distance between adjacent passers-by is quantified through the calculation formula of the centroid distance index, and compared with the preset centroid distance threshold interval: if the distance is within the normal range, it is judged as normal passage; if the distance is abnormally close, it is automatically marked as "adjacent passers-by with abnormal distance". This quantitative analysis avoids the traditional tailing judgment based on subjective experience, and ensures the scientificity and objectivity of tailing behavior detection. This step judges the relative position relationship of all people in real time by performing pairwise adjacent analysis on the grayscale areas of all passers-by in the passage area: the system can quickly screen In complex traffic scenarios (such as rush hour or fast traffic), the system can still accurately identify abnormal tailgating situations and improve the robustness of the system. By calculating the centroid coordinates and centroid spacing index of each personnel area, it provides accurate data input for subsequent comprehensive analysis of tailing anomalies: the centroid coordinates, as the core parameter of personnel position, support spacing calculation, speed detection and other analyses. The centroid spacing index, as one of the important indicators of tailing behavior, quantifies the abnormal proximity between personnel. This data support ensures the overall logical coherence of the tailgating detection system and the accuracy of the analysis results. The system can automatically determine whether the centroid spacing index exceeds the threshold and mark the abnormal situation, reducing human intervention and judgment time, realizing the automated processing of tailgating detection, and improving the execution efficiency and real-time performance of the system.

[0034] Specifically, the specific formula for calculating the abnormal comprehensive index of tailgating of adjacent passers-by with abnormal spacing in a set passage area is as follows: Among them, WsY is the comprehensive index of tailing abnormality of adjacent travelers with abnormal spacing in the set passage area, ZxJ is the center of mass spacing index of adjacent travelers with abnormal spacing in the set passage area, μ1 is the spacing coefficient stored in the database, ScY is the speed difference index of adjacent travelers with abnormal spacing in the set passage area, μ2 is the speed coefficient stored in the database, LjP is the path angle deviation index of adjacent travelers with abnormal spacing in the set passage area, μ3 is the angle deviation coefficient stored in the database, LmZ is the contour area index of adjacent travelers with abnormal spacing in the set passage area, μ4 is the contour area coefficient stored in the database, μ1+μ2+μ3+μ4=1, π is pi, and the value is 3.14 in this embodiment.

[0035] In this implementation scheme, by constructing a comprehensive index formula for tailing anomalies, multiple key indicators (center of mass distance index, speed difference index, path angle deviation index, contour area index) and their corresponding weight coefficients are integrated and calculated to achieve a quantitative evaluation of the tailing behavior. This comprehensive quantitative analysis method can judge the severity of the tailing behavior more comprehensively and objectively, making up for the deficiency of single parameter judgment in the prior art. Four core parameters are introduced in the formula: center of mass distance index (quantifying the actual spatial proximity between personnel), speed difference index (analyzing the change in movement speed of the front and rear personnel and identifying abnormal acceleration phenomena), path angle deviation index (reflecting the consistency of the movement direction of the two people, the tailing behavior often has a highly consistent path), contour area index (combining the size of the personnel area to judge the rationality of the relative position and distance). These parameters complement each other and can comprehensively capture the spatial characteristics, movement characteristics and contour characteristics of the tailing behavior, effectively reduce the false positive rate and missed positive rate, and improve the detection accuracy and reliability of the system. The spacing coefficient, speed coefficient, angle deviation coefficient and contour area coefficient in the formula are stored in the database and can be dynamically adjusted and optimized according to the actual traffic scene and historical detection data: the weight coefficient is continuously optimized through historical data, so that the system can maintain high detection performance in different scenes; the empirical data stored in the database is used as the basis for judgment, so that the tailgating anomaly comprehensive index is more scientific and reasonable. The introduction of pi in the formula makes the tailgating anomaly comprehensive index have higher numerical discrimination when calculating, effectively avoiding the phenomenon of convergence of index values ​​under different parameter combinations, and further improving the ability to distinguish the severity of tailgating behavior in a refined manner. After the tailgating anomaly comprehensive index is calculated, it can be compared with the tailing anomaly comprehensive index threshold: if the index value exceeds the threshold, the system determines that there is tailgating behavior and takes anti-tailgating measures; if the index value is lower than the threshold, normal passage is allowed. This quantitative threshold judgment method can flexibly set the threshold standard according to the actual detection data to ensure the flexibility and adaptability of the system judgment logic.

[0036] Specifically, when there are adjacent travelers with abnormal spacing in a set passage area, the specific steps of analyzing the speed difference index and path angle deviation index of the adjacent travelers with abnormal spacing in the set passage area are as follows: for two travelers among the adjacent travelers with abnormal spacing in the grayscale image of the passage area, the centroid coordinates of several time points are continuously analyzed respectively, and a comprehensive analysis is performed to obtain the travel speed values ​​and speed vectors of the two travelers among the adjacent travelers with abnormal spacing in the grayscale image of the passage area; and the travel speed values ​​and speed vectors of the two travelers among the adjacent travelers with abnormal spacing in the grayscale image of the passage area are comprehensively analyzed to obtain the speed difference index and path angle deviation index of the adjacent travelers with abnormal spacing in the set passage area.

[0037] The specific formula for calculating the speed difference index and path angle deviation index of adjacent pedestrians with abnormal spacing in the set passage area is as follows: Among them, ScY is the speed difference index of adjacent travelers with abnormal spacing in the set passage area, e is a natural explanation, and in this embodiment, the value is 2.718, SdZ1 is the travel speed value of the first traveler among the adjacent travelers with abnormal spacing in the grayscale image of the passage area, SdZ2 is the travel speed value of the second traveler among the adjacent travelers with abnormal spacing in the grayscale image of the passage area, LjP is the path angle deviation index of the adjacent travelers with abnormal spacing in the set passage area, is the velocity vector of the first passerby among the adjacent passersby with abnormal spacing in the grayscale image of the passing area, It is the velocity vector of the second passerby among the adjacent passersby with abnormal spacing in the grayscale image of the passage area.

[0038] In this implementation scheme, by continuously analyzing the centroid coordinates of several time points, the travel speed value and speed vector of the passers-by can be calculated in real time, and the movement state of the person can be dynamically captured. This method converts static image data into dynamic motion features, thereby more comprehensively analyzing the behavior of the person, and effectively solves the problem that it is difficult to reflect the movement state of the person in real time in the prior art. By calculating the speed difference index, the speed change between adjacent passers-by is quantified: if the speed difference is small, it indicates that the person behind has the possibility of "following" the person in front; if the speed difference is large, it may be a normal passage or a fast passing phenomenon. This quantitative analysis method ensures the scientificity and objectivity of the tailing behavior detection, and can accurately distinguish the tailing phenomenon from the normal passage behavior, reduce the misjudgment rate, and calculate the path angle deviation index to analyze the angle between the speed vectors of the two people, and judge the consistency of the movement directions of the two people: a small path angle deviation indicates that the movement directions of the two people are highly consistent, and the possibility of tailing is high; a large path angle deviation indicates that the directions of the two people are obviously different , the possibility of tailgating is low. This method of quantifying the path direction can effectively capture the behavior of the tailer trying to imitate the path of the person passing in front, improve the system's comprehensive recognition ability of the tailing phenomenon, and introduce natural constants in the speed difference index formula, so that the index calculation result has an exponential level of discrimination, which enhances the detection sensitivity under different speed difference conditions: the smaller the speed difference, the larger the index value, which helps to quickly identify the tailing behavior; the larger the speed difference, the smaller the index value, which helps to exclude normal traffic conditions. This design optimizes the system's processing ability for edge data and makes the recognition result of the tailing behavior more refined. Combining the speed difference index with the path angle deviation index, a multi-dimensional dynamic analysis of the movement status of the person is realized: the speed feature reflects the change in the movement speed of the person passing by; the direction feature reflects the consistency of the movement path of the person passing by. By comprehensively analyzing these two indicators, the system can more comprehensively capture the essential characteristics of the tailing behavior, make up for the shortcomings of the single parameter analysis, and greatly improve the accuracy and robustness of the tailing behavior detection.

[0039] Specifically, when there are adjacent passers-by with abnormal spacing in the set passage area, the specific steps of analyzing the contour area index of the adjacent passers-by with abnormal spacing in the set passage area are as follows: the number of grayscale pixel points of the passers-by in the first and second grayscale areas of the adjacent passers-by with abnormal spacing in the grayscale image of the passage area are counted respectively, and they are marked as the contour area values ​​of the first and second grayscale areas of the adjacent passers-by with abnormal spacing in the grayscale image of the passage area, and a comprehensive analysis is performed to obtain the contour area index of the adjacent passers-by with abnormal spacing in the grayscale image of the passage area, and the calculation formula is as follows: Among them, LmZ is the contour area index of adjacent passers-by with abnormal spacing in the set passage area, LkM1 is the contour area value of the grayscale area of ​​the first passer-by among the adjacent passers-by with abnormal spacing in the grayscale image of the passage area, and LkM2 is the contour area value of the grayscale area of ​​the second passer-by among the adjacent passers-by with abnormal spacing in the grayscale image of the passage area.

[0040] In this implementation scheme, by calculating the contour area index, the difference in the size of the grayscale areas of adjacent passers-by can be quantified, directly reflecting the relative area occupied by the two persons in the image: closer contour areas indicate that the two persons are at a similar distance and the possibility of tailgating is higher; larger differences in contour areas indicate that the person behind may be farther away and the possibility of tailing is lower. This quantitative analysis method based on contour area effectively solves the defect that traditional methods cannot distinguish the distances between adjacent persons. The contour area index is based on the statistics and comparison of the number of grayscale pixels, and the calculation logic is simple and efficient: no complex geometric model or image reconstruction is required, only the number of pixels in the connected domain needs to be counted. The calculation formula directly performs a ratio analysis on the contour area values ​​of the two passers-by, ensuring real-time and rapid response capabilities. This design enables the system to maintain efficient and accurate tailgating detection performance in high traffic flow scenarios. The contour area not only reflects the size of the area occupied by passers-by, but also indirectly analyzes the distance and status characteristics of the person's position: a larger contour indicates that the person is closer to the camera. Or in an abnormally close state, a smaller outline indicates that the person is in a rear position of normal traffic. By comprehensively analyzing the ratio of the outline areas of the two people, the system can more accurately judge the tailing behavior and reduce the misjudgment rate. The outline area index, as one of the core parameters of tailing detection, can be fused and analyzed with other indicators (such as center of mass distance index, speed difference index, path angle deviation index) to form a comprehensive index of tailing anomaly. The combination of multiple parameters enables the system to make a comprehensive judgment on tailing behavior from multiple dimensions such as spatial relationship, motion state, and contour features. The outline area index fills the loopholes that may be caused by relying solely on position or motion parameters, and enhances the judgment accuracy and robustness of the system. In actual applications, factors such as the distance, body shape, and posture of the passers-by will affect the size of the outline area in the image. By real-time counting the outline areas of adjacent passers-by and calculating their ratio, the system can quickly mark the suspicious state of adjacent passers-by with abnormal spacing. This process provides accurate data input for subsequent comprehensive analysis of tailing, ensuring the consistency and scientificity of the overall judgment logic of the system.

[0041] Specifically, Figure 3As shown, the specific steps for obtaining the threshold value of the comprehensive index of tailgating anomaly are as follows: real-time statistics of the time intervals of several people passing through the anti-tailgating gate are made, and a comprehensive analysis is performed to obtain the average time interval; the time interval parameter value and the initial comprehensive index threshold of tailgating anomaly are obtained, and a comprehensive analysis is performed in combination with the average time interval to obtain the comprehensive index threshold of tailgating anomaly.

[0042] The specific formulas for calculating the mean time interval and the threshold of the comprehensive index of trailing anomalies are as follows: Among them, TxJ is the mean time interval of the passage, XjG s is the time interval value of the sth person passing through the anti-tailgating gate, ZhY is the tailgating abnormality comprehensive index threshold, ChY is the initial tailing abnormality comprehensive index threshold, α is the adjustment coefficient stored in the database, TxC is the time interval parameter value, s=1,2,3,…,s0,s0 is the number of people passing through the anti-tailgating gate.

[0043] In this implementation scheme, by real-time statistics of the time interval values ​​of the passage through the anti-tailgating gate and calculating the mean time interval, the system can dynamically adjust the tailgating abnormality comprehensive index threshold according to the actual passage status. This method effectively solves the problem of poor adaptability of fixed thresholds in different scenarios, ensuring that the system can maintain accurate tailgating detection performance under different traffic conditions such as peak and off-peak periods. The passage time interval parameter value, the initial tailgating abnormality comprehensive index threshold and the adjustment coefficient are introduced. Combined with the real-time calculated time interval mean, a new tailgating abnormality comprehensive index threshold is calculated through comprehensive analysis. This method has the following advantages: the threshold is dynamically optimized according to the actual passage data, and the system is more in line with the actual passage scene; by combining historical passage data with real-time monitoring data, the scientificity and rationality of threshold calculation are improved. Fixed thresholds are prone to misjudgment or missed judgment in complex passage environments, and the dynamically adjusted tailgating abnormality comprehensive index threshold can be optimized according to the actual distribution of passage time intervals: a shorter time interval indicates an increased possibility of tailgating behavior, the threshold is automatically lowered, and the system is more sensitive; a longer time interval indicates a higher probability of normal passage, the threshold is automatically increased to reduce misjudgment. This dynamic adjustment mechanism enables tailgating The detection is more accurate, which effectively reduces the system false triggering caused by unreasonable thresholds. In different time periods, the traffic flow will change. The system reflects the current traffic status by counting the real-time traffic time interval: for high-traffic scenarios, the traffic time interval is short, and the system automatically lowers the tailgating threshold to ensure that tailgaters cannot pass through; for low-traffic scenarios, the traffic time interval is long, and the system appropriately raises the threshold to avoid misjudging normal passers-by. This dynamic adaptive design for traffic changes effectively improves the robustness and practical application effect of the system. The design is based on the easy-to-obtain parameter of the traffic time interval. Through simple statistics and adjustment calculations, it is easy to integrate and deploy in the existing system. At the same time, the system can flexibly adjust the threshold through the adjustment coefficient and the database storage data to adapt to different scenarios and needs. It has strong scalability and compatibility. The calculation of the mean traffic time interval and the tailing anomaly comprehensive index threshold provides accurate data support for system decision-making: the mean traffic time interval reflects the average traffic pattern of passers-by; the dynamic threshold ensures that the system makes flexible tailgating detection judgments based on the real-time status. This data support design enables the system to achieve more intelligent and accurate tailgating behavior detection.

[0044] In summary, this application has at least the following effects:

[0045] By acquiring image data of the passage area in real time and performing grayscale and binarization processing, the system can efficiently extract the passage area. Combined with the connected domain analysis and centroid coordinate calculation, the system can accurately locate the center coordinates of the passage personnel, and then judge whether there are adjacent passage personnel with abnormal spacing through the centroid spacing index. Compared with the simple static image judgment of the existing technology, this method can dynamically monitor the passage status, effectively identify tailing behavior in complex scenarios, reduce the misjudgment and missed judgment rates, and improve the accuracy and reliability of tailing detection.

[0046] By calculating the centroid distance index, speed difference index, path angle deviation index and contour area index of adjacent travelers with abnormal spacing, the spatial position, movement speed, travel direction and regional contour characteristics between the people are comprehensively analyzed. The quantitative analysis of these parameters ensures that the system can fully capture the multi-dimensional characteristics of tailgating behavior. Through the comprehensive evaluation of the comprehensive index of tailing anomaly, the system can more accurately distinguish between tailing phenomena and normal traffic behavior, solving the problem of misjudgment of a single indicator.

[0047] By real-time statistics of the average time intervals of the anti-tailgating gate, combined with the preset time interval parameter value and the initial threshold, the tailgating detection standard is dynamically optimized. This method effectively adapts to different traffic scenarios (such as high traffic and low traffic) and improves the robustness and flexibility of the system. The dynamic threshold adjustment not only reduces the misjudgment of normal passers-by, but also enhances the practicality and traffic efficiency of the system in actual applications.

[0048] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0049] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An anti-tailgating gate system based on image recognition, characterized in that: include: Data acquisition module, identification analysis module, judgment analysis module, index analysis module, comprehensive analysis module, anti-tailing judgment module, anti-tailing execution module; The data acquisition module is used to acquire the traffic area image data of the set traffic area in real time, and the traffic area image data includes a number of pixel points in the traffic area image, and the pixel value and two-dimensional coordinates of each pixel point; The recognition and analysis module is used to recognize and analyze the traffic area image data of the set traffic area in real time, and determine whether there is a traffic area for people in the set traffic area according to the recognition and analysis results; The judgment and analysis module is used to judge in real time whether there are adjacent passing persons with abnormal spacing in the set passage area when there is a passing person area in the set passage area; The index analysis module is used to analyze the centroid distance index, speed difference index, path angle deviation index, and contour area index of adjacent passers-by with abnormal spacing in the set passage area when there are adjacent passers-by with abnormal spacing in the set passage area; The comprehensive analysis module is used to comprehensively analyze the centroid distance index, speed difference index, path angle deviation index, and contour area index of adjacent passers-by with abnormal spacing in a set passage area to obtain a comprehensive index of tailing abnormality of adjacent passers-by with abnormal spacing in the set passage area; The anti-tailgating judgment module is used to judge whether the tailgating abnormality comprehensive index of adjacent pedestrians with abnormal spacing is higher than the tailgating abnormality comprehensive index threshold; The anti-tailgating execution module is used to input a preset anti-tailgating measure to the anti-tailgating gate when the tailgating abnormality comprehensive index of adjacent travelers with abnormal spacing is higher than the tailgating abnormality comprehensive index threshold.

2. The anti-tailgating gate system based on image recognition according to claim 1 is characterized in that: The specific steps for judging whether there is a pedestrian zone in the set passage area based on the recognition analysis results are as follows: Grayscale processing is performed on the passage area image data of the set passage area to obtain the passage area grayscale image data of the set passage area, wherein the passage area grayscale image data includes a plurality of grayscale pixel points in the passage area grayscale image and a grayscale pixel value of each grayscale pixel point; Binarization is performed on the grayscale image data of the set passage area based on a preset grayscale pixel interval to obtain a number of grayscale pixel points of passers-by in the grayscale image of the passage area; If the number of grayscale pixels of the passers-by in the grayscale image of the pass area is lower than the preset grayscale pixel number threshold, it indicates that there is no passer-by area in the set pass area; If the number of grayscale pixels of passers-by in the grayscale image of the passage area is greater than or equal to a preset grayscale pixel number threshold, it indicates that there is a passer-by area in the set passage area.

3. The anti-tailgating gate system based on image recognition according to claim 2 is characterized in that: The specific steps for real-time determination of whether there are adjacent pedestrians with abnormal spacing in the set passage area are as follows: Performing connected domain analysis on a plurality of grayscale pixel points of passers-by in the grayscale image of the pass area, obtaining a plurality of grayscale regions of passers-by in the grayscale image of the pass area, and each grayscale region of passers-by includes a plurality of grayscale pixel points of passers-by; For each grayscale region of a passerby in the grayscale image of the passerby region, the two-dimensional coordinates of each grayscale pixel point of each passerby in the grayscale region of the passerby are read respectively and a comprehensive analysis is performed to obtain the centroid coordinates of each grayscale region of the passerby in the grayscale image of the passerby region; Performing adjacent centroid distance analysis on each grayscale area of ​​a passerby in the grayscale image of the pass area and each remaining grayscale area of ​​a passerby in the grayscale image of the pass area, respectively, to obtain centroid distance indexes of several groups of adjacent passersby in the set pass area; Determine whether the centroid distance index of each group of adjacent pedestrians in the set passage area is within a preset centroid distance threshold range; If each centroid distance index in the set passage area is within the preset centroid distance threshold range, it means that there are no adjacent passers-by with abnormal distance in the set passage area; If there are adjacent passers-by in the set passage area whose centroid distance index is outside the preset centroid distance threshold range, they are marked as adjacent passers-by with abnormal distance; The specific formula for calculating the centroid coordinates of each grayscale area of ​​a passer-by in the grayscale image of the pass area is as follows: Among them, (x′ i , y′ i ) is the centroid coordinate of the grayscale area of ​​the i-th passer-by in the grayscale image of the pass-by area, (x ir ,y ir ) is the two-dimensional coordinate of the rth grayscale pixel point of the i-th grayscale area of ​​the passerby in the grayscale image of the passerby area, i = 1, 2, 3, ..., i0, i0 is the number of grayscale areas of the passerby in the grayscale image of the passerby area.

4. The anti-tailgating gate system based on image recognition according to claim 3 is characterized in that: in, The specific formula for calculating the centroid distance index of each group of adjacent passers-by in the set passage area is as follows: Among them, ZxJ t is the centroid distance index of the tth group of adjacent passers-by in the set passage area, (x′ t1 , y′ t1 ) is the centroid coordinate of the grayscale area of ​​the first passerby in the tth group of adjacent passersby in the grayscale image of the pass area, (x′ t2 , y′ t2 ) is the centroid coordinate of the grayscale area of ​​the second passerby in the tth group of adjacent passersby in the grayscale image of the passage area, t = 1, 2, 3, ..., t0, t0 is the number of adjacent passerby groups in the grayscale image of the passage area.

5. The anti-tailgating gate system based on image recognition according to claim 1, characterized in that: The specific formula for calculating the comprehensive index of tailgating anomaly of adjacent passers-by with abnormal spacing in a set passage area is as follows: Among them, WsY is the comprehensive index of tailing abnormality of adjacent travelers with abnormal spacing in the set passage area, ZxJ is the center of mass spacing index of adjacent travelers with abnormal spacing in the set passage area, μ1 is the spacing coefficient stored in the database, ScY is the speed difference index of adjacent travelers with abnormal spacing in the set passage area, μ2 is the speed coefficient stored in the database, LjP is the path angle deviation index of adjacent travelers with abnormal spacing in the set passage area, μ3 is the angle deviation coefficient stored in the database, LmZ is the contour area index of adjacent travelers with abnormal spacing in the set passage area, μ4 is the contour area coefficient stored in the database, μ1+μ2+μ3+μ4=1, and π is pi.

6. The anti-tailgating gate system based on image recognition according to claim 1, characterized in that: When there are adjacent passers-by with abnormal spacing in the set passage area, the specific steps of analyzing the speed difference index and path angle deviation index of adjacent passers-by with abnormal spacing in the set passage area are as follows: For two passers-by among the adjacent passers-by with abnormal spacing in the grayscale image of the passage area, the centroid coordinates of several time points are continuously analyzed respectively, and a comprehensive analysis is performed to obtain the travel speed values ​​and speed vectors of the two passers-by among the adjacent passers-by with abnormal spacing in the grayscale image of the passage area; A comprehensive analysis is then performed on the travel speed values ​​and speed vectors of two adjacent travelers with abnormal spacing in the grayscale image of the passage area to obtain the speed difference index and path angle deviation index of adjacent travelers with abnormal spacing in the set passage area.

7. The anti-tailgating gate system based on image recognition according to claim 6, characterized in that: The specific formula for calculating the speed difference index and path angle deviation index of adjacent pedestrians with abnormal spacing in the set passage area is as follows: Among them, ScY is the speed difference index of adjacent travelers with abnormal spacing in the set passage area, e is a natural explanation, SdZ1 is the travel speed value of the first traveler among the adjacent travelers with abnormal spacing in the grayscale image of the passage area, SdZ2 is the travel speed value of the second traveler among the adjacent travelers with abnormal spacing in the grayscale image of the passage area, LjP is the path angle deviation index of adjacent travelers with abnormal spacing in the set passage area, is the velocity vector of the first passerby among the adjacent passersby with abnormal spacing in the grayscale image of the passing area, It is the velocity vector of the second passerby among the adjacent passersby with abnormal spacing in the grayscale image of the passage area.

8. The anti-tailgating gate system based on image recognition according to claim 1, characterized in that: When there are adjacent passers-by with abnormal spacing in the set passage area, the specific steps of analyzing the contour area index of adjacent passers-by with abnormal spacing in the set passage area are as follows: The number of grayscale pixels of the first and second passers-by in the grayscale image of the passage area with abnormal spacing between adjacent passers-by is counted respectively, and marked as the contour area values ​​of the first and second passers-by in the grayscale image of the passage area with abnormal spacing between adjacent passers-by, respectively. After comprehensive analysis, the contour area index of the abnormal spacing between adjacent passers-by in the grayscale image of the passage area is obtained, and the calculation formula is as follows: Among them, LmZ is the contour area index of adjacent passers-by with abnormal spacing in the set passage area, LkM1 is the contour area value of the grayscale area of ​​the first passer-by among the adjacent passers-by with abnormal spacing in the grayscale image of the passage area, and LkM2 is the contour area value of the grayscale area of ​​the second passer-by among the adjacent passers-by with abnormal spacing in the grayscale image of the passage area.

9. The anti-tailgating gate system based on image recognition according to claim 1, characterized in that: The specific steps to obtain the threshold of the comprehensive index of trailing anomalies are as follows: Real-time statistics of the time intervals of several people passing through the anti-tailgating gate are collected, and comprehensive analysis is performed to obtain the average time interval; Obtain the travel time interval parameter value and the initial tailing anomaly comprehensive index threshold, and perform a comprehensive analysis based on the travel time interval mean value to obtain the tailing anomaly comprehensive index threshold.

10. The anti-tailgating gate system based on image recognition according to claim 9, characterized in that: The specific formulas for calculating the mean time interval and the threshold of the comprehensive index of trailing anomalies are as follows: Among them, TxJ is the mean time interval of the passage, XjG s is the time interval value of the sth person passing through the anti-tailgating gate, ZhY is the tailgating abnormality comprehensive index threshold, ChY is the initial tailing abnormality comprehensive index threshold, α is the adjustment coefficient stored in the database, TxC is the travel time interval parameter value, s=1, 2, 3, …, s0, s0 is the number of people passing through the anti-tailgating gate.

Images