Monitoring method and system based on equatorial plane projection

Abstract

The present invention relates to the field of image monitoring technology, specifically to a monitoring method and system based on equatorial projection, comprising the following steps: based on a target monitoring area, converting the position of a target monitoring object in video data into spherical coordinates centered on a pan-tilt camera, converting the spherical coordinate data into equatorial coordinate data where the pan-tilt camera is located, and obtaining two-dimensional equatorial projection image data. The present invention captures video images in real time through a pan-tilt camera, and converts the position of a target monitoring object in the image into spherical coordinates and then performs equatorial projection, so that the scene observed at any viewing angle is converted into an absolute position on the equatorial plane, maintaining the invariance of the scene position, facilitating subsequent analysis and monitoring, and by comparing the deviation of the moving speed and duration of the target monitoring object from a preset standard, as well as the matching degree analysis of the behavior pattern and the abnormal behavior pattern database, it is possible to accurately identify and evaluate the abnormal behavior risk of the monitoring object, thereby implementing accurate early warning measures.

Description

Technical Field

[0001] The present invention relates to the field of image monitoring technology, and in particular to a monitoring method and system based on equatorial plane projection. Background Art

[0002] The field of image monitoring technology covers various technical methods and systems that use video equipment to monitor the environment or specific places in real time or non-real time. Aiming to monitor and record the environment and activities through various forms of video capture devices, image monitoring technology is widely used in public safety, traffic monitoring, industrial automation, home safety, and commercial security. With the advancement of technology, the field of image monitoring has developed from simple video recording to intelligent functions including motion detection, facial recognition, and behavior analysis, which has greatly improved monitoring efficiency and accuracy.

[0003] Monitoring methods are a series of methods that use technical means to observe and record specific areas, objects or individuals. The methods usually include setting the position and angle of the camera, controlling the movement of the camera, and processing and analyzing the captured video data. Monitoring methods are widely used, mainly for security protection, such as monitoring illegal activities and ensuring the safety of people in public places, residential areas, commercial facilities, etc. In addition, monitoring methods are also used in traffic monitoring, industrial production process monitoring and behavioral research to improve operational efficiency and safety management level.

[0004] Traditional monitoring methods face the challenges of complex data processing and long response time when processing and analyzing monitoring data. Traditional methods rely on fixed or preset camera angles, which leads to blind spots or inaccurate data capture in actual operations once the monitoring angle needs to be adjusted. Traditional methods lack highly flexible and adaptive monitoring mechanisms in terms of object movement recognition and behavior pattern analysis, making it difficult to accurately evaluate and respond to abnormal behaviors in real time, resulting in security monitoring loopholes. For example, in public security monitoring, due to the inability to effectively identify or predict abnormal behavior patterns, traditional methods are unable to detect or prevent security incidents in a timely manner, affecting the monitoring effect and the overall performance of security protection. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings in the prior art and to propose a monitoring method and system based on equatorial plane projection.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a monitoring method based on equatorial plane projection comprises the following steps:

[0007] S1: Based on the target monitoring area, the video image data of the monitoring area is captured in real time through the PTZ camera, the position of the target monitoring object in the video data is converted into spherical coordinates centered on the PTZ camera, spherical coordinate data is obtained, and the spherical coordinate data is converted into equatorial plane coordinate data where the PTZ camera is located, and two-dimensional equatorial projection image data is obtained;

[0008] S2: Based on the two-dimensional equatorial projection image data, analyzing the moving speed and moving duration of the target monitoring object, calculating the deviation of the target monitoring object from the normal moving speed and moving duration, evaluating the moving deviation index, and obtaining the moving deviation information;

[0009] S3: Based on the two-dimensional equatorial projection image data, analyzing the movement path and behavior pattern of the target monitoring object, identifying the abnormal behavior pattern of the target monitoring object, and calculating the abnormal behavior risk index, evaluating the abnormal behavior pattern risk, and obtaining the abnormal behavior identification result;

[0010] S4: Based on the movement deviation information and abnormal behavior recognition results, the overall abnormal risk level of the target object is calculated. According to the abnormal risk level, corresponding early warning measures are implemented, and the angle of the gimbal camera is adjusted to continuously monitor the target monitoring object information to obtain safety early warning detection information.

[0011] The present invention is improved in that the steps of obtaining the spherical coordinate data are:

[0012] S111: Based on the target monitoring area, the video image data of the monitoring area is captured in real time by the PTZ camera, and the pixel coordinates of the target object in the image are identified;

[0013] S112: Based on the pixel coordinates of the target object in the image, by the formula:

[0014]

[0015] Convert pixel coordinates into spherical coordinates centered on the gimbal camera to obtain spherical coordinate data;

[0016] Where θ is the latitude angle in spherical coordinates, φ is the longitude angle in spherical coordinates, x and y are the pixel coordinates of the target object, x0 and y0 are the pixel coordinates of the center of the image, and f x and f y are the focal lengths of the camera in the x and y directions, d x and d y is the optical distortion adjustment factor.

[0017] The present invention is improved in that the steps of acquiring the two-dimensional equatorial projection image data are:

[0018] S121: extracting the latitude angle θ and the longitude angle φ in the spherical coordinates based on the spherical coordinate data;

[0019] S122: Based on the latitude angle θ and the longitude angle φ, the formula:

[0020]

[0021] Calculate the equatorial coordinate data, convert the spherical coordinate data into the equatorial coordinate data of the gimbal camera, and obtain the two-dimensional equatorial projection image data;

[0022] Where R is the radius of the sphere, α is the inclination angle between the PTZ camera and the ground, θ is the latitude angle in spherical coordinates, φ is the longitude angle in spherical coordinates, and x eq and eq It is the two-dimensional coordinate of the target object on the equatorial projection plane.

[0023] The present invention is improved in that the step of obtaining the movement deviation information is:

[0024] S211: Based on the two-dimensional equatorial projection image data, extract the position change of the target monitoring object between two consecutive frames to obtain the actual moving speed v act and moving time t act ;

[0025] S212: Based on the actual moving speed v act and moving time t act , with the preset normal moving speed v norm and moving time t norm Compare and calculate the speed deviation Δv and duration deviation Δt, and use the formula:

[0026]

[0027] Calculate the moving deviation index P, evaluate the moving deviation degree according to the value of the moving deviation index P, and obtain the moving deviation information;

[0028] Among them, v act is the actual moving speed, t act is the actual moving time, v norm and t norm are the preset normal moving speed and moving duration, Δv and Δt represent the deviation of speed and duration, respectively, and w v and w t are the weight coefficients of speed and duration deviation respectively, and P is the movement deviation index.

[0029] The present invention is improved in that the method for identifying the abnormal behavior pattern of the target monitoring object is:

[0030] S311: Based on the two-dimensional equatorial projection image data, extract the moving path information of the target monitoring object, including the position coordinates of the target in the continuous time frame, through the formula:

[0031]

[0032] Calculate the matching score M between the current behavior pattern of the target monitored object and the abnormal behavior pattern;

[0033] Among them, w i is the weight of the i-th point, λ i is the sensitivity factor, τ is the tolerance threshold, d i is the distance between the actual path point and the preset pattern point, M is the matching score, n is the total number of data points, and e is the base of the natural logarithm;

[0034] S312: Based on the matching score M, the matching score M is compared with a preset matching threshold, and the behavior where M exceeds the matching threshold is marked as abnormal behavior, thereby obtaining the abnormal behavior type of the target object.

[0035] The present invention is improved in that the step of obtaining the abnormal behavior recognition result is:

[0036] S321: Based on the abnormal behavior type of the target object, extract the matching score M corresponding to the abnormal behavior type j ;

[0037] S322: Based on the matching score M j , through the formula:

[0038]

[0039] Calculate the abnormal behavior risk index R, evaluate the abnormal behavior risk of the target object according to the value of the abnormal behavior risk index R, and obtain the abnormal behavior identification result;

[0040] Where k is the total number of matched abnormal behavior patterns, M j is the matching score of the jth behavior pattern, w j is the risk weight of the jth behavior pattern, α j is the sensitivity coefficient of the jth behavior mode, τ j is the threshold of the jth behavior pattern, and R is the abnormal behavior risk index.

[0041] The present invention is improved in that the method for implementing the corresponding early warning measures is:

[0042] S411: Based on the movement deviation information and the abnormal behavior recognition result, the movement deviation index P and the abnormal behavior risk index R are extracted, using the formula:

[0043]

[0044] Calculate the overall abnormal risk level E of the target object;

[0045] Among them, P is the moving deviation index, R is the abnormal behavior risk index, γ is the weight coefficient for adjusting the impact of moving deviation, δ is the weight coefficient for adjusting the impact of behavioral risk, ∈ is the overall weight coefficient, and E is the overall abnormal risk level;

[0046] S412: Based on the overall abnormal risk level E, the overall abnormal risk level E is compared with a preset abnormal standard, and corresponding early warning measures are implemented.

[0047] A monitoring system based on equatorial projection, the system comprising:

[0048] The video data capture module captures real-time video image data through a pan-tilt camera based on the target monitoring area, converts the position of the target monitoring object in the video data into spherical coordinates centered on the pan-tilt camera, and generates a real-time spherical coordinate set;

[0049] The equatorial projection conversion module should convert the spherical coordinate data into the equatorial coordinate data of the PTZ camera based on the real-time spherical coordinate set, and obtain the equatorial projection image data;

[0050] The behavior speed analysis module analyzes the movement speed and duration of the monitored object based on the equatorial plane projection image data, compares it with the preset standard, calculates the deviation, and generates a movement deviation index;

[0051] The abnormal behavior pattern matching module compares the behavior pattern of the monitored object with a preset abnormal behavior pattern database based on the equatorial plane projection image data, identifies the abnormal behavior type, and calculates the abnormal behavior risk index to obtain abnormal behavior pattern risk information;

[0052] The abnormal behavior risk assessment module calculates the overall abnormal risk level based on the movement deviation index and abnormal behavior pattern risk information, implements matching early warning measures, and adjusts the pan-tilt camera angle to continuously monitor the target monitoring object information to obtain safety early warning detection information.

[0053] Compared with the prior art, the advantages and positive effects of the present invention are:

[0054] In the present invention, video images are captured in real time by a pan-tilt camera, and the position of the target monitored object in the image is converted into spherical coordinates and then projected onto the equatorial plane, so that the scenes observed at any viewing angle are converted into absolute positions on the equatorial plane, maintaining the invariance of the scene positions, facilitating subsequent analysis and monitoring, enhancing the stability and accuracy of monitoring, reducing data redundancy and processing complexity, and accurately identifying and evaluating the abnormal behavior risks of the monitored object by comparing the deviations of the moving speed and duration of the target monitored object from preset standards, as well as analyzing the matching degree of the behavior pattern with the abnormal behavior pattern database, thereby implementing accurate early warning measures, optimizing the accuracy and real-time performance of the behavior analysis, and enabling the monitoring system to more effectively prevent and manage potential security risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 is a flow chart of the method of the present invention;

[0056] Figure 2 A flow chart for obtaining spherical coordinate data of the present invention;

[0057] Figure 3 A flow chart for obtaining two-dimensional equatorial projection image data according to the present invention;

[0058] Figure 4 A flow chart of obtaining movement deviation information of the present invention;

[0059] Figure 5 A flow chart of the present invention for identifying abnormal behavior patterns of target monitoring objects;

[0060] Figure 6 A flow chart of obtaining abnormal behavior recognition results for the present invention;

[0061] Figure 7 The present invention is a flowchart of implementing corresponding early warning measures. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0063] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0064] Example

[0065] See also Figure 1 The present invention provides a technical solution: a monitoring method based on equatorial plane projection, comprising the following steps:

[0066] S1: Based on the target monitoring area, the video image data of the monitoring area is captured in real time through the PTZ camera, the position of the target monitoring object in the video data is converted into spherical coordinates centered on the PTZ camera, spherical coordinate data is obtained, and the spherical coordinate data is converted into equatorial plane coordinate data where the PTZ camera is located, and two-dimensional equatorial projection image data is obtained;

[0067] S2: Based on the two-dimensional equatorial projection image data, the moving speed and moving duration of the target monitoring object are analyzed, and compared with the preset normal moving speed and moving duration, the deviation between the target monitoring object and the normal moving speed and moving duration is calculated, the moving deviation index is evaluated, and the moving deviation information is obtained;

[0068] S3: Based on the two-dimensional equatorial projection image data, analyze the moving path and behavior pattern of the target monitoring object, compare the moving path of the target monitoring object with the preset abnormal behavior pattern database, calculate the matching degree between the current behavior pattern of the target monitoring object and the abnormal behavior pattern, identify the abnormal behavior pattern of the target monitoring object, and calculate the abnormal behavior risk index, evaluate the risk of the abnormal behavior pattern, and obtain the abnormal behavior identification result;

[0069] S4: Based on the movement deviation information and abnormal behavior recognition results, the overall abnormal risk level of the target object is calculated. According to the abnormal risk level, corresponding early warning measures are implemented, and the angle of the pan-tilt camera is adjusted to continuously monitor the target monitoring object information to obtain safety early warning detection information;

[0070] The two-dimensional equatorial projection image data includes the two-dimensional position coordinates and corresponding timestamp of the target monitored object. The movement deviation information includes the change in movement speed and movement duration of the target monitored object. The abnormal behavior recognition results include the risk score and abnormal type of abnormal behavior. The safety warning detection information includes the corresponding warning measures and pan-tilt camera adjustment information.

[0071] See also Figure 2 , the steps to obtain spherical coordinate data are:

[0072] S111: Based on the target monitoring area, the video image data of the monitoring area is captured in real time by the PTZ camera, and the pixel coordinates of the target object in the image are identified;

[0073] S112: Based on the pixel coordinates of the target object in the image, the formula is:

[0074]

[0075] Convert pixel coordinates into spherical coordinates centered on the gimbal camera to obtain spherical coordinate data;

[0076] Where θ is the latitude angle in spherical coordinates, φ is the longitude angle in spherical coordinates, x and y are the pixel coordinates of the target object, x0 and y0 are the pixel coordinates of the center of the image, and f x and f y are the focal lengths of the camera in the x and y directions, d x and d y is the optical distortion adjustment factor.

[0077] formula:

[0078]

[0079] Parameter details and how to obtain them:

[0080] Pixel coordinates x and y: Coordinates obtained after identifying the target object from the image captured by the camera through image recognition technology. Indicates the location of the target object in the image. These coordinates can be extracted directly from the video stream in real time through image processing software or algorithms, such as the image recognition function in the OpenCV library.

[0081] Image center coordinates x0 and y0: Usually half of the image resolution. For example, if you are using a 1920x1080 HD camera, x0 is 960 and y0 is 540. This coordinate is used to determine the center point of the image and is an inherent property of the image resolution.

[0082] Focal length f x and f y: The parameters are usually provided by the camera manufacturer or can be obtained through the optical calibration process. It represents the focal length of the camera on the x-axis and y-axis, which is part of the camera's optical characteristics.

[0083] Optical distortion adjustment coefficient d x and d y : Coefficients are used to correct for distortions in pixel positions due to lens curvature and other optical distortions. The coefficients can be obtained through an optical calibration procedure, such as taking multiple shots with a chessboard pattern and analyzing the results to determine the values, or they can be preset by an expert.

[0084] Calculation example:

[0085] Parameters and assumed values: Pixel coordinates of the target object x=1000 and y=800; Image center coordinate x0

[0086] =960, y0=540 focal length; f x = 1000 pixels, f y =1000 pixels; distortion adjustment factor d x =1.05,d y =0.95.

[0087] Calculate the latitude angle θ:

[0088]

[0089] Substituting the given values:

[0090]

[0091] The formula used to calculate the longitude angle φ is:

[0092]

[0093] The calculated results θ≈15.5° and φ≈2.18° indicate that, in the viewing angle captured by the gimbal camera, the target object is located approximately 15.5 degrees above the image center and approximately 2.18 degrees to the right.

[0094] See also Figure 3 , the steps to obtain the two-dimensional equatorial projection image data are:

[0095] S121: extracting the latitude angle θ and the longitude angle φ in the spherical coordinates based on the spherical coordinate data;

[0096] S122: Based on the latitude angle θ and the longitude angle φ, the formula:

[0097]

[0098] Calculate the equatorial coordinate data, convert the spherical coordinate data into the equatorial coordinate data of the gimbal camera, and obtain the two-dimensional equatorial projection image data;

[0099] Where R is the radius of the sphere, α is the inclination angle between the PTZ camera and the ground, θ is the latitude angle in spherical coordinates, φ is the longitude angle in spherical coordinates, and x eq and eq It is the two-dimensional coordinate of the target object on the equatorial projection plane.

[0100] formula:

[0101]

[0102] Parameter details and how to obtain them:

[0103] Latitude angle θ and longitude angle φ: obtained by the previous steps.

[0104] Sphere radius R: In most monitoring applications, the radius of the sphere can be assumed to be unity to simplify calculations and for normalization in calculations and conversions.

[0105] The tilt angle α between the PTZ camera and the ground is an actual measurement value, which is determined by the physical tilt angle of the camera during installation. It can be obtained through the specifications provided by the installation engineer or directly measured using a tilt sensor.

[0106] Calculation example:

[0107] Assumptions:

[0108] θ=30°(latitude angle)

[0109] φ=45°(longitude angle)

[0110] R = 1 (radius of the sphere, normalized)

[0111] α=10°(Camera tilt angle)

[0112] Conversion process and calculation formula:

[0113] Convert θ and φ to radians for calculation:

[0114]

[0115] Use the formula to calculate the equatorial coordinates:

[0116]

[0117] Specific calculation:

[0118]

[0119]

[0120] therefore

[0121] The calculation results of the equatorial coordinates are:

[0122]

[0123] The calculated x eq and eq The coordinates represent the position of the target object on the equatorial plane.

[0124] See also Figure 4 , the steps to obtain the movement deviation information are:

[0125] S211: Based on the two-dimensional equatorial projection image data, extract the position change of the target monitoring object between two consecutive frames to obtain the actual moving speed v act and moving time t act ;

[0126] S212: Based on the actual moving speed v act and moving time t act , with the preset normal moving speed v norm and moving time t norm Compare and calculate the speed deviation Δv and duration deviation Δt, and use the formula:

[0127]

[0128] Calculate the moving deviation index P, evaluate the moving deviation degree according to the value of the moving deviation index P, and obtain the moving deviation information;

[0129] Among them, v act is the actual moving speed, t act is the actual moving time, v norm and t norm are the preset normal moving speed and moving duration, Δv and Δt represent the deviation of speed and duration, respectively, and w v and w t are the weight coefficients of speed and duration deviation respectively, and P is the movement deviation index.

[0130] formula:

[0131]

[0132] Parameter details and how to obtain them:

[0133] Actual moving speed v act : It is obtained by measuring the position change of the target between two consecutive frames from the two-dimensional equatorial projection image data and dividing it by the time difference.act It is calculated by measuring the pixel distance of the target between two frames and converting it into actual distance, then dividing it by the time interval between the two frames.

[0134] Actual moving time t act : Determined by recording the time between when the target starts moving and when it stops moving. Usually obtained directly from the timestamp of the video surveillance system.

[0135] Default normal movement speed v norm and moving time t norm : Parameters are set based on previously observed or desired behavior. Parameters are set by the system administrator or through historical data analysis.

[0136] Speed ​​deviation Δv and duration deviation Δt: Speed ​​deviation is calculated by Δv=v act -v norm The calculated time deviation is obtained by Δt=t act -t norm Calculated. Deviation shows how different the actual observations are from the expected standard.

[0137] Weight w v and w t : The weight coefficient is pre-set according to the specific needs of the monitoring scenario. It reflects the relative importance of speed and time deviation when evaluating the total deviation index.

[0138] Calculation example:

[0139] Assume the following parameters:

[0140] v act =5 m / s, t act = 120 seconds, v norm =4 m / s (default normal speed), t norm = 100 seconds (default normal duration), w v =0.6, w t =0.4.

[0141] Calculation process:

[0142] Calculate speed deviation and duration deviation:

[0143] Δv = 5-4 = 1 m / s

[0144] Δt=120-100=20 seconds

[0145] Calculate the deviation index P:

[0146]

[0147] The obtained deviation index P≈0.332 indicates that the target's movement speed and duration show slight abnormalities relative to the preset standard.

[0148] See also Figure 5 , the method of identifying abnormal behavior patterns of target monitoring objects is:

[0149] S311: Based on the two-dimensional equatorial projection image data, the moving path information of the target monitoring object is extracted, including the position coordinates of the target in the continuous time frame, through the formula:

[0150]

[0151] Calculate the matching score M between the current behavior pattern of the target monitored object and the abnormal behavior pattern;

[0152] Among them, w i is the weight of the i-th point, λ i is the sensitivity factor, τ is the tolerance threshold, d i is the distance between the actual path point and the preset pattern point, M is the matching score, n is the total number of data points, and e is the base of the natural logarithm;

[0153] S312: Based on the matching score M, the matching score M is compared with a preset matching threshold, and the behavior where M exceeds the matching threshold is marked as abnormal behavior, thereby obtaining the abnormal behavior type of the target object.

[0154] formula:

[0155]

[0156] Parameter details and how to obtain them:

[0157] d i : The distance between the actual path point and the preset mode point, calculated as: d i =|D i -Q i |, where D i represents the actual observed position of the target monitoring object at the i-th time point in the image. The position is directly extracted from the monitoring video through image processing technology. i is the expected position of the ith time point in the preset abnormal behavior pattern database. The data is preset based on previous abnormal behavior cases or theoretical models.

[0158] w i : The weight of each point, which is preset according to the importance of the point in the behavior pattern. For example, the key turning points of the behavior pattern will be given higher weights.

[0159] λ i: A sensitivity factor for each point, used to adjust the influence of distance deviation. A high sensitivity factor means that even a small deviation will result in a large score difference.

[0160] τ: Tolerance threshold, used to define what degree of deviation is acceptable. Deviations exceeding this threshold will significantly reduce the matching score.

[0161] Calculation example:

[0162] Assume the following parameters and points:

[0163] P={(2,3),(3,4),(5,5)}actual path points

[0164] Q={(2,2),(3,5),(5,6)}Preset mode points

[0165] w={0.5,1,0.5}weight

[0166] λ = {1,1,1} sensitivity factor

[0167] τ=1 tolerance threshold

[0168] Calculation process:

[0169] Calculate the distance deviation d for each point i :

[0170] d1=|(2-2,3-2)|=1

[0171] d2=|(3-3,4-5)|=1

[0172] d3=|(5-5,5-6)|=1

[0173] Application matching calculation formula:

[0174]

[0175] The calculated matching score M=1.0 indicates that the actually observed path has a high consistency with the preset abnormal behavior pattern, that is, the monitored object exhibits the preset abnormal behavior type.

[0176] See also Figure 6 ,The steps to obtain abnormal behavior recognition results are:

[0177] S321: Based on the abnormal behavior type of the target object, extract the matching score M corresponding to the abnormal behavior type j ;

[0178] S322: Based on the matching score M j , through the formula:

[0179]

[0180] Calculate the abnormal behavior risk index R, evaluate the abnormal behavior risk of the target object according to the value of the abnormal behavior risk index R, and obtain the abnormal behavior identification result;

[0181] Where k is the total number of matched abnormal behavior patterns, M j is the matching score of the jth behavior pattern, w j is the risk weight of the jth behavior pattern, α j is the sensitivity coefficient of the jth behavior mode, τ j is the threshold of the jth behavior pattern, and R is the abnormal behavior risk index.

[0182] Parameter details and how to obtain them:

[0183] Matching score M j : Indicates the degree of match between the behavior of the target monitored object and the preset abnormal behavior pattern, which is calculated through the previous steps.

[0184] Weight w j : Indicates the weight corresponding to the risk level of each behavior pattern. The weight is preset based on historical data, expert opinions, or the severity of the consequences of the behavior. The purpose is to emphasize behavior patterns that are more likely to lead to serious consequences in risk assessment.

[0185] Sensitivity coefficient α j : Controls the influence of the match score on the risk assessment. A high coefficient value means that even a slightly higher match score will significantly increase the risk assessment, usually adjusted for the specific risk sensitivity of the pattern.

[0186] Threshold τ j : is the minimum threshold of the matching score for judging a behavior pattern as abnormal. If the matching score of a behavior pattern is lower than this value, its contribution to the risk index is small.

[0187] Calculation example:

[0188] Assume there are three behavior modes, and the corresponding parameter settings are as follows:

[0189] Matching score M = {0.8, 0.5, 0.9}

[0190] Weight w = {0.5, 1.0, 0.7}

[0191] Sensitivity coefficient α = {1.5, 1.0, 2.0}

[0192] Threshold τ = {0.6, 0.5, 0.7}

[0193] Calculation process:

[0194] The risk contribution of each mode is calculated by the following formula:

[0195]

[0196] For each mode, calculate:

[0197] For the first pattern:

[0198]

[0199] For the second pattern:

[0200]

[0201] For the third mode:

[0202]

[0203] Calculate the abnormal behavior risk index R:

[0204] R=R1+R2+R3

[0205] =0.183+0.125+0.339

[0206] =0.647

[0207] The calculation result R=0.647 indicates that according to the matching of the current behavior data with the preset abnormal pattern, the overall behavior pattern of the target monitored object shows a medium to high abnormal behavior risk.

[0208] See also Figure 7 , the methods for implementing corresponding early warning measures are:

[0209] S411: Based on the movement deviation information and abnormal behavior recognition results, the movement deviation index P and the abnormal behavior risk index R are extracted by the formula:

[0210]

[0211] Calculate the overall abnormal risk level E of the target object;

[0212] Among them, P is the moving deviation index, R is the abnormal behavior risk index, γ is the weight coefficient for adjusting the impact of moving deviation, δ is the weight coefficient for adjusting the impact of behavioral risk, ∈ is the overall weight coefficient, and E is the overall abnormal risk level;

[0213] S412: Based on the overall abnormal risk level E, the overall abnormal risk level E is compared with the preset abnormal standard, and corresponding early warning measures are implemented.

[0214] formula:

[0215]

[0216] Parameter details and how to obtain them:

[0217] Movement deviation index P: calculated by the previous step.

[0218] Abnormal behavior risk index R: calculated through the previous steps.

[0219] Weight coefficient γ: used to adjust the influence of the movement deviation index P in the overall risk level calculation. The weight is usually set by the system administrator based on specific monitoring requirements and historical data, reflecting the importance of abnormal movement speed in the overall risk.

[0220] Weight coefficient δ: It is used to adjust the influence of abnormal behavior risk index R in the overall risk level calculation. The weight is set based on monitoring requirements and historical data, reflecting the importance of abnormal behavior patterns in the overall risk.

[0221] Weight coefficient ∈: is used to adjust the influence of the sum of movement deviation and abnormal behavior risk index on the overall risk level. The weight is used to balance the relationship between movement deviation and behavioral risk to ensure that the combined impact of the two is properly considered.

[0222] Calculation example:

[0223] Assumptions:

[0224] Moving deviation index P = 0.4

[0225] Abnormal behavior risk index R = 0.6

[0226] Weight coefficient γ = 0.3

[0227] Weight coefficient δ = 0.4

[0228] Weight coefficient ∈ = 0.3

[0229] Calculation process:

[0230] Calculate the contribution of each part:

[0231] γ·P 2 =0.3 · 0.4 2 =0.3 0.16 = 0.048

[0232] δ·R 2 =0.4·0.6 2 =0.4 0.36 = 0.144

[0233] ∈·log(1+P+R)=0.3·log(1+0.4+0.6)

[0234] =0.3·log(2)

[0235] ≈0.3 0.3010

[0236] =0.0903

[0237] Calculate the overall abnormal risk level E:

[0238]

[0239] The calculated result E≈0.5314 indicates that the overall abnormal risk level of the target is medium. This value provides a quantitative risk assessment result, which can be used to determine whether early warning measures or further monitoring adjustments are needed.

[0240] The monitoring system based on equatorial projection includes:

[0241] The video data capture module captures real-time video image data through a pan-tilt camera based on the target monitoring area, converts the position of the target monitoring object in the video data into spherical coordinates centered on the pan-tilt camera, and generates a real-time spherical coordinate set;

[0242] The equatorial projection conversion module is based on the real-time spherical coordinate set, and should convert the spherical coordinate data to the equatorial coordinate data where the PTZ camera is located, and obtain the equatorial projection image data;

[0243] The behavior speed analysis module analyzes the movement speed and duration of the monitored object based on the equatorial projection image data, compares it with the preset standard, calculates the deviation, and generates a movement deviation index;

[0244] The abnormal behavior pattern matching module compares the behavior pattern of the monitored object with the preset abnormal behavior pattern database based on the equatorial plane projection image data, identifies the abnormal behavior type, and calculates the abnormal behavior risk index to obtain the abnormal behavior pattern risk information;

[0245] The abnormal behavior risk assessment module calculates the overall abnormal risk level based on the movement deviation index and abnormal behavior pattern risk information, implements matching early warning measures, and adjusts the gimbal camera angle to continuously monitor the target monitoring object information and obtain safety early warning detection information.

[0246] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. The monitoring method based on equatorial projection is characterized in that: The following steps are involved: Based on the target monitoring area, the video image data of the monitoring area is captured in real time through the pan-tilt camera, the position of the target monitoring object in the video data is converted into spherical coordinates centered on the pan-tilt camera, spherical coordinate data is obtained, and the spherical coordinate data is converted into equatorial plane coordinate data where the pan-tilt camera is located, and two-dimensional equatorial projection image data is obtained; Based on the two-dimensional equatorial projection image data, analyzing the moving speed and moving duration of the target monitoring object, calculating the deviation of the target monitoring object from the normal moving speed and moving duration, evaluating the moving deviation index, and obtaining the moving deviation information; The steps for obtaining the movement deviation information are as follows: Based on the two-dimensional equatorial projection image data, the position change of the target monitoring object between two consecutive frames is extracted to obtain the actual moving speed v act and moving time t act ; Based on the actual moving speed v act and moving time t act , with the preset normal moving speed v norm and moving time t norm Compare and calculate the speed deviation Δv and duration deviation Δt, and use the formula: Calculate the moving deviation index P, evaluate the moving deviation degree according to the value of the moving deviation index P, and obtain the moving deviation information; Among them, v act is the actual moving speed, t act is the actual moving time, v norm and t norm are the preset normal moving speed and moving duration, Δv and Δt represent the deviation of speed and duration, respectively, and w v and w t are the weight coefficients of speed and duration deviation, respectively, and P is the movement deviation index; Based on the two-dimensional equatorial projection image data, analyzing the movement path and behavior pattern of the target monitoring object, identifying the abnormal behavior pattern of the target monitoring object, and calculating the abnormal behavior risk index, evaluating the abnormal behavior pattern risk, and obtaining the abnormal behavior identification result; Based on the movement deviation information and abnormal behavior recognition results, the overall abnormal risk level of the target object is calculated. According to the abnormal risk level, corresponding early warning measures are implemented, and the angle of the pan-tilt camera is adjusted to continuously monitor the target monitoring object information to obtain safety early warning detection information.

2. The monitoring method based on equatorial plane projection according to claim 1, characterized in that: The steps for obtaining the spherical coordinate data are: Based on the target monitoring area, the PTZ camera captures the video image data of the monitoring area in real time and identifies the pixel coordinates of the target object in the image; Based on the pixel coordinates of the target object in the image, by the formula: Convert pixel coordinates into spherical coordinates centered on the gimbal camera to obtain spherical coordinate data; Where θ is the latitude angle in spherical coordinates, φ is the longitude angle in spherical coordinates, x and y are the pixel coordinates of the target object, x0 and y0 are the pixel coordinates of the center of the image, and f x and f y are the focal lengths of the camera in the x and y directions, d x and d y is the optical distortion adjustment factor.

3. The monitoring method based on equatorial projection according to claim 2, characterized in that: The steps for acquiring the two-dimensional equatorial projection image data are: Based on the spherical coordinate data, extract the latitude angle θ and the longitude angle φ in the spherical coordinates; Based on the latitude angle θ and longitude angle φ, by the formula: Calculate the equatorial coordinate data, convert the spherical coordinate data into the equatorial coordinate data of the gimbal camera, and obtain the two-dimensional equatorial projection image data; Where R is the radius of the sphere, α is the inclination angle between the PTZ camera and the ground, θ is the latitude angle in spherical coordinates, φ is the longitude angle in spherical coordinates, and x eq and eq It is the two-dimensional coordinate of the target object on the equatorial projection plane.

4. The monitoring method based on equatorial projection according to claim 1, characterized in that: The method for identifying the abnormal behavior pattern of the target monitoring object is: Based on the two-dimensional equatorial projection image data, the moving path information of the target monitoring object is extracted, including the position coordinates of the target in the continuous time frame, through the formula: Calculate the matching score M between the current behavior pattern of the target monitored object and the abnormal behavior pattern; Among them, w i is the weight of the i-th point, λ i is the sensitivity factor, τ is the tolerance threshold, d i is the distance between the actual path point and the preset pattern point, M is the matching score, n is the total number of data points, and e is the base of the natural logarithm; Based on the matching score M, the matching score M is compared with a preset matching threshold, and the behavior in which M exceeds the matching threshold is marked as abnormal behavior, thereby obtaining the abnormal behavior type of the target object.

5. The monitoring method based on equatorial projection according to claim 4, characterized in that: The steps for obtaining the abnormal behavior recognition result are: Based on the abnormal behavior type of the target object, extract the matching score M corresponding to the abnormal behavior type j ; Based on the matching score M j , through the formula: Calculate the abnormal behavior risk index R, evaluate the abnormal behavior risk of the target object according to the value of the abnormal behavior risk index R, and obtain the abnormal behavior identification result; Where k is the total number of matched abnormal behavior patterns, M j is the matching score of the jth behavior pattern, w j is the risk weight of the jth behavior pattern, α j is the sensitivity coefficient of the jth behavior mode, τ j is the threshold of the jth behavior pattern, and R is the abnormal behavior risk index.

6. The monitoring method based on equatorial projection according to claim 1, characterized in that: The method for implementing the corresponding early warning measures is: Based on the movement deviation information and abnormal behavior recognition results, the movement deviation index P and the abnormal behavior risk index R are extracted, through the formula: Calculate the overall abnormal risk level E of the target object; Among them, P is the moving deviation index, R is the abnormal behavior risk index, γ is the weight coefficient for adjusting the impact of moving deviation, δ is the weight coefficient for adjusting the impact of behavioral risk, ∈ is the overall weight coefficient, and E is the overall abnormal risk level; Based on the overall abnormal risk level E, the overall abnormal risk level E is compared with a preset abnormal standard, and corresponding early warning measures are implemented.

7. The monitoring system based on equatorial projection is characterized by: According to the monitoring method based on equatorial plane projection according to any one of claims 1 to 6, the system comprises: The video data capture module captures real-time video image data through a pan-tilt camera based on the target monitoring area, converts the position of the target monitoring object in the video data into spherical coordinates centered on the pan-tilt camera, and generates a real-time spherical coordinate set; The equatorial projection conversion module should convert the spherical coordinate data into the equatorial coordinate data of the PTZ camera based on the real-time spherical coordinate set, and obtain the equatorial projection image data; The behavior speed analysis module analyzes the movement speed and duration of the monitored object based on the equatorial plane projection image data, compares it with the preset standard, calculates the deviation, and generates a movement deviation index; The method for calculating the moving deviation index is: Based on the two-dimensional equatorial projection image data, the position change of the target monitoring object between two consecutive frames is extracted to obtain the actual moving speed v act and moving time t act ; Based on the actual moving speed v act and moving time t act , with the preset normal moving speed v norm and moving time t norm Compare and calculate the speed deviation Δv and duration deviation Δt, and use the formula: Calculate the moving deviation index P; Among them, v act is the actual moving speed, t act is the actual moving time, v norm and t norm are the preset normal moving speed and moving duration, Δv and Δt represent the deviation of speed and duration, respectively, and w v and w t are the weight coefficients of speed and duration deviation, respectively, and P is the movement deviation index; The abnormal behavior pattern matching module compares the behavior pattern of the monitored object with a preset abnormal behavior pattern database based on the equatorial plane projection image data, identifies the abnormal behavior type, and calculates the abnormal behavior risk index to obtain abnormal behavior pattern risk information; The abnormal behavior risk assessment module calculates the overall abnormal risk level based on the movement deviation index and abnormal behavior pattern risk information, implements matching early warning measures, and adjusts the pan-tilt camera angle to continuously monitor the target monitoring object information to obtain safety early warning detection information.

Images