Railway trestle video intelligent analysis and application method based on oil depot self-control signal triggering

Abstract

The invention provides a railway trestle video intelligent analysis application method based on oil depot self-control signal triggering, and the method comprises the steps: trying to obtain an oil depot self-control signal in an oil depot railway oil distribution operation process; if the attempt is successful, a target camera of the trestle platform is triggered to shoot the trestle platform, and a railway trestle video is obtained; based on an intelligent video analysis technology, performing off-duty behavior analysis on the guardian according to the railway trestle video to obtain an off-duty behavior analysis result; and based on an off-duty behavior analysis result, outputting a visual result and performing early warning. According to the railway trestle video intelligent analysis application method based on oil depot self-control signal triggering, intelligent video analysis is triggered based on the recognition result of the oil depot self-control signal, the off-duty behavior analysis result of the guardian is obtained, the occurrence of the guardian off-duty phenomenon is avoided, the intelligence and refinement of oil depot management are improved, the safety risk is reduced, and the working efficiency is improved. And the management level is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of target detection, and in particular to an application method for intelligent analysis of railway trestle video based on oil depot automatic control signal triggering. Background Art

[0002] Railway oil delivery is a key link in the operation process of oil depots, and its importance is self-evident. This link involves the efficient and safe transportation and storage of flammable and explosive materials, which has a decisive impact on the safe and stable operation of the entire oil depot. In this process, supervisors, as the direct persons responsible for on-site safety management, bear vital responsibilities. They must not only ensure the standardization of the operation process, but also remain vigilant at all times to prevent potential safety risks.

[0003] However, in actual operations, the phenomenon of absenteeism frequently occurs. Absenteeism means that the supervisor fails to stick to his post and leaves the operation site without authorization, resulting in the loss of effective monitoring of the operation process. This behavior not only violates the basic principles of safety management, but also poses a serious threat to the safe and stable operation of the oil depot.

[0004] In view of this, there is an urgent need for an intelligent analysis application method of railway trestle video based on oil depot automatic control signal triggering to at least solve the above-mentioned deficiencies. Summary of the invention

[0005] One of the purposes of the present invention is to provide an application method for intelligent analysis of railway trestle video based on oil depot automatic control signal triggering. Based on the recognition result of the oil depot automatic control signal, intelligent video analysis is triggered to obtain the analysis result of the off-duty behavior of the supervisor, thereby realizing real-time target recognition and tracking of the entire process of railway oil dispatching operation of the oil depot, avoiding the occurrence of supervisor off-duty phenomenon, improving the intelligence and refinement of oil depot management, reducing safety risks, and improving management level.

[0006] The railway trestle video intelligent analysis application method based on oil depot automatic control signal triggering provided by the embodiment of the present invention includes:

[0007] Try to obtain the oil depot automatic control signal during the oil depot railway oil delivery operation;

[0008] If the acquisition attempt is successful, multiple target cameras on the trestle platform are triggered to shoot the trestle platform to obtain the railway trestle video;

[0009] Based on intelligent video analysis technology, the off-duty behavior of supervisors is analyzed based on the railway trestle video to obtain the off-duty behavior analysis results;

[0010] Based on the analysis results of off-duty behavior, visual results are output and warnings are issued.

[0011] Preferably, it is characterized in that attempting to obtain the oil depot automatic control signal during the oil depot railway oil delivery operation includes:

[0012] Establish a communication link with the communication node of the host computer in the central control room;

[0013] Based on the communication link, the oil depot automatic control signal is obtained.

[0014] Preferably, if the acquisition attempt is successful, multiple target cameras on the trestle platform are triggered to shoot the trestle platform to obtain the railway trestle video, including:

[0015] If the acquisition attempt is successful, obtain the control template of the target camera;

[0016] Based on the control template, the target camera is controlled to shoot the trestle platform to obtain the railway trestle video.

[0017] Preferably, based on intelligent video analysis technology, the off-post behavior analysis of the guardian is performed according to the railway trestle video, and the off-post behavior analysis results are obtained, including:

[0018] Based on YOLO v8, target detection and target tracking are performed on the supervisors in the railway trestle video to obtain detection and tracking results;

[0019] Conduct off-duty behavior analysis based on the detection and tracking results to obtain the off-duty behavior analysis results.

[0020] Preferably, performing off-duty behavior analysis based on the detection and tracking results to obtain off-duty behavior analysis results includes:

[0021] According to the detection and tracking results, detecting whether there is a corresponding supervisor in each first working area of ​​the trestle platform;

[0022] If there is no corresponding guardian in the first working area, the corresponding first working area is used as the second working area, and the corresponding non-existent guardian is used as the first tracking person;

[0023] Obtaining the duration of the unidentified state of the first tracked person in the corresponding second working area;

[0024] If the duration of the unrecognized state is less than or equal to the preset threshold of the duration of the unrecognized state, an off-duty behavior analysis is performed on the first tracked person to obtain a first off-duty behavior analysis sub-result;

[0025] If the duration of the unrecognized state is longer than the preset threshold of the duration of the unrecognized state, the second off-duty behavior analysis sub-result of the first tracked person is output;

[0026] The first off-duty behavior analysis sub-result and the second off-duty behavior analysis sub-result are integrated to obtain the off-duty behavior analysis result.

[0027] Preferably, if the duration of the unrecognized state is less than or equal to a preset threshold of the duration of the unrecognized state, an off-duty behavior analysis is performed on the first tracked person to obtain a first off-duty behavior analysis sub-result, including:

[0028] Attempt to detect a backup guardian in a secondary work area;

[0029] If the substitute guardian attempts to obtain the information successfully, the first off-duty behavior analysis sub-result of the first tracking person in the second work area is output as not off-duty;

[0030] If the substitute guardian fails in his attempt to obtain it, the corresponding second working area will be used as the third working area;

[0031] Try to obtain the temporary leave request of the first tracking person corresponding to the third work area;

[0032] If the temporary leave request is successfully obtained, the first off-duty behavior analysis sub-result of the first tracked person in the third work area is output as not off-duty;

[0033] If the temporary leave request fails to be obtained, the corresponding third work area is used as the fourth work area, and at the same time, the first tracking person corresponding to the fourth work area is used as the second tracking person;

[0034] Obtain the shooting blind area of ​​the fourth working area;

[0035] Determine the strategy for acquiring blind spot information based on the blind spots in shooting;

[0036] Obtain blind spot information according to a blind spot information acquisition strategy;

[0037] Determine whether the second tracking person is in the corresponding fourth working area according to the blind area information;

[0038] If the second tracked person is in the corresponding fourth work area, outputting the first off-duty behavior analysis sub-result of the second tracked person in the corresponding fourth work area as not off-duty;

[0039] If the second tracked person is not in the corresponding fourth work area, the first off-duty behavior analysis sub-result of the second tracked person in the corresponding fourth work area is output as off-duty.

[0040] Preferably, attempting to detect a substitute guardian in the second working area includes:

[0041] Detecting the first tracking person in the second working area other than the first tracking person corresponding to the second working area as the third tracking person;

[0042] Obtaining the previous communication information between the third tracking person and the first tracking person corresponding to the second working area;

[0043] If the previous communication information meets the substitute communication conditions, and there is a fourth tracker in the second work area that the third tracker is responsible for, the third tracker is determined to be the substitute guardian in the corresponding second work area;

[0044] Among them, the conditions for substitute exchange include:

[0045] The preceding communication action includes at least one first confirmation action by a third tracking person;

[0046] and,

[0047] After the first confirmation action, at least a second confirmation action of a fourth tracking person can be obtained; the time interval between the first confirmation action and the second confirmation action is less than or equal to a preset time interval threshold;

[0048] and,

[0049] Pre-order communication semantics includes the semantics of pre-set time representation and the semantics of substitute event representation.

[0050] Preferably, according to the shooting blind area, a blind area information acquisition strategy is determined, including:

[0051] Acquire scanning point cloud data of the fourth working area;

[0052] Constructing a scanning point cloud model of the fourth working area according to the scanning point cloud data;

[0053] Based on the current target camera parameters, set the virtual camera in the scanned point cloud model;

[0054] Determine the shooting blind area in the scanned point cloud model;

[0055] Simulate shooting of the blind area with a virtual camera to obtain a simulated shooting result;

[0056] If the simulated shooting result is successful, determine the simulated shooting parameters of the virtual camera;

[0057] Perform blind spot shooting according to the virtual camera simulation shooting parameters and target camera parameters;

[0058] If the simulated shooting result is a shooting failure, a verification reminder is sent to the second tracking person in the corresponding fourth working area.

[0059] The railway trestle video intelligent analysis application system based on oil depot automatic control signal triggering provided by the embodiment of the present invention comprises:

[0060] The oil depot automatic control signal identification subsystem is used to try to obtain the oil depot automatic control signal during the oil depot railway oil dispatching operation;

[0061] The railway trestle video shooting subsystem is used to trigger multiple target cameras on the trestle platform to shoot the trestle platform and obtain the railway trestle video if the acquisition attempt is successful;

[0062] The off-post behavior analysis subsystem is used to analyze the off-post behavior of the supervisor based on the railway trestle video based on the intelligent video analysis technology and obtain the off-post behavior analysis results;

[0063] The analysis result warning subsystem is used to output visual results and issue warnings based on the off-duty behavior analysis results.

[0064] Preferably, the oil depot automatic control signal identification subsystem includes:

[0065] A communication link establishment module is used to establish a communication link with the communication node of the host computer in the central control room;

[0066] The oil depot automatic control signal acquisition module is used to obtain the oil depot automatic control signal based on the communication link.

[0067] Preferably, the railway trestle video shooting subsystem includes:

[0068] A control template acquisition module is used to obtain the control template of the target camera if the acquisition attempt is successful;

[0069] The railway trestle video acquisition module is used to control the target camera to shoot the trestle platform based on the control template and acquire the railway trestle video.

[0070] The beneficial effects of the present invention are:

[0071] The present invention triggers intelligent video analysis based on the recognition result of the oil depot's automatic control signal, obtains the analysis result of the guardian's off-duty behavior, realizes real-time target recognition and tracking of the entire process of railway oil dispatching operations in the oil depot, avoids the occurrence of guardian off-duty phenomenon, improves the intelligence and refinement of oil depot management, reduces safety risks, and improves the management level.

[0072] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in the present application documents.

[0073] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0075] Figure 1 It is a schematic diagram of an application method of railway trestle video intelligent analysis based on oil depot automatic control signal triggering in an embodiment of the present invention;

[0076] Figure 2 It is a schematic diagram of the railway trestle video intelligent analysis application system based on the oil depot automatic control signal triggering in an embodiment of the present invention. DETAILED DESCRIPTION

[0077] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0078] The embodiment of the present invention provides a railway trestle video intelligent analysis application method based on oil depot automatic control signal triggering, such as Figure 1 As shown, including:

[0079] Try to obtain the oil depot automatic control signal during the oil depot railway oil delivery operation;

[0080] If the acquisition attempt is successful, multiple target cameras on the trestle platform are triggered to shoot the trestle platform to obtain the railway trestle video;

[0081] Based on intelligent video analysis technology, the off-duty behavior of supervisors is analyzed based on the railway trestle video to obtain the off-duty behavior analysis results;

[0082] Based on the analysis results of off-duty behavior, visual results are output and warnings are issued.

[0083] The working principle and beneficial effects of the above technical solution are:

[0084] The oil depot automatic control signal is sent by the host computer when the oil loading starts. When the railway sends oil to the car for quantitative loading, the operator in the central control room first sets the oil delivery parameters on the host computer, including the oil delivery method, the number of liters of oil delivered, etc. After confirmation, click the start button to start the railway oil delivery. After the server receives the oil depot automatic control signal, it uses multiple target cameras on the trestle platform to obtain the railway trestle video during the oil delivery period. The off-post behavior analysis is to enable the intelligent video analysis function, identify and monitor the personnel on the trestle platform, and continuously detect whether the trestle platform remains occupied. The off-post behavior analysis results are: determine whether the trestle platform is The results of judging whether the staff in the working area of ​​the station are off-duty, for example: if the result of intelligent video analysis is that there are people continuously, then the result of off-duty behavior analysis is that they are not off-duty; if the result of intelligent video analysis is that there are no people during period A, but there are obstacles blocking the view, and the result of intelligent video analysis is that there are people during period B, then the result of off-duty behavior analysis is that they are not off-duty; and if the continuous detection result during period C is that there is no one, then the result of off-duty behavior analysis is that they are off-duty; the visualization results include: video clips and data of the guardians who are judged to be off-duty, including time, oil depot, location, screenshots and other information; warnings include: pop-up windows for management personnel and linkage voice warnings.

[0085] The present invention triggers intelligent video analysis based on the recognition result of the oil depot's automatic control signal, obtains the analysis result of the guardian's off-duty behavior, realizes real-time target recognition and tracking of the entire process of railway oil dispatching operations in the oil depot, avoids the occurrence of guardian off-duty phenomenon, improves the intelligence and refinement of oil depot management, reduces safety risks, and improves the management level.

[0086] In one embodiment, an attempt is made to obtain an oil depot automatic control signal during an oil depot railway oil delivery operation, including:

[0087] Establish a communication link with the communication node of the host computer in the central control room;

[0088] Based on the communication link, the oil depot automatic control signal is obtained.

[0089] The working principle and beneficial effects of the above technical solution are:

[0090] The present invention establishes network communication with the upper computer of the central control room to obtain the oil depot automatic control signal, and the oil depot automatic control signal acquisition is more reasonable.

[0091] In one embodiment, if the acquisition attempt is successful, multiple target cameras on the trestle platform are triggered to shoot the trestle platform to obtain the railway trestle video, including:

[0092] If the acquisition attempt is successful, the control template of the target camera is obtained; wherein the control template includes a plurality of control instructions predefined for the target camera, such as: a command to turn on a certain camera, a command to turn off a certain camera;

[0093] Based on the control template, the target camera is controlled to shoot the trestle platform to obtain the railway trestle video.

[0094] The working principle and beneficial effects of the above technical solution are:

[0095] The present invention introduces a control template to obtain the railway trestle video shot by a target camera, thereby improving the automation level of triggering the target camera.

[0096] In one embodiment, based on intelligent video analysis technology, the off-duty behavior of the supervisor is analyzed according to the railway trestle video, and the off-duty behavior analysis results are obtained, including:

[0097] Based on YOLO v8, target detection and target tracking are performed on the supervisors in the railway trestle video to obtain detection and tracking results;

[0098] Conduct off-duty behavior analysis based on the detection and tracking results to obtain the off-duty behavior analysis results.

[0099] The working principle and beneficial effects of the above technical solution are:

[0100] The present invention introduces the YOLO v8 model to perform target detection and target tracking of the supervisor in the railway trestle video, obtains the detection and tracking results and performs behavior analysis, thereby ensuring the high precision of the recognition process and the high accuracy of the recognition results. Since YOLO v8 has the characteristics of multi-scale feature fusion and strong robustness, it performs better in dealing with problems such as long trestle platforms, large near targets and small far targets, many platform equipment that are easily blocked, complex backgrounds, and light changes, thereby improving the accuracy of the off-duty behavior analysis results.

[0101] In one embodiment, the off-duty behavior analysis is performed according to the detection and tracking results, and the off-duty behavior analysis results are obtained, including:

[0102] According to the detection and tracking results, it is detected whether there is a corresponding supervisor in each first working area of ​​the trestle platform; wherein the first working area is: an area on the trestle platform that is pre-set and requires supervisors to work, and the corresponding supervisor in the corresponding supervisor means that the supervisor's work attributes are consistent with the work attributes of the first working area;

[0103] If there is no corresponding guardian in the first working area, the corresponding first working area is used as the second working area, and the corresponding non-existent guardian is used as the first tracking person;

[0104] Obtaining the duration of the unrecognized state of the first tracking person in the corresponding second working area; wherein the duration of the unrecognized state is: the duration of the state in which the first tracking person is not recognized or confirmed by the system in the monitoring area;

[0105] If the duration of the unrecognized state is less than or equal to the preset duration threshold of the unrecognized state, an off-duty behavior analysis is performed on the first tracked person to obtain a first off-duty behavior analysis sub-result; wherein the preset duration threshold of the unrecognized state is preset manually; the first off-duty behavior analysis sub-result is: the result obtained by performing an off-duty behavior analysis on the first tracked person when the duration of the unrecognized state is less than or equal to the duration threshold of the unrecognized state;

[0106] If the duration of the unrecognized state is longer than the preset duration threshold of the unrecognized state, output the second off-duty behavior analysis sub-result of the first tracked person; the second off-duty behavior analysis sub-result is: the result of the off-duty behavior analysis of the first tracked person when the duration of the unrecognized state is longer than the threshold of the duration of the unrecognized state;

[0107] The first off-duty behavior analysis sub-result and the second off-duty behavior analysis sub-result are integrated to obtain the off-duty behavior analysis result.

[0108] The working principle and beneficial effects of the above technical solution are:

[0109] When conducting off-duty analysis, there may be situations where the supervisor is blocked by obstacles or leaves temporarily or under reasonable circumstances. Therefore, when there is no corresponding supervisor in the first work area, a specific analysis is required.

[0110] First, determine the first tracking person, and obtain the duration of the unidentified state in the corresponding second work area; compare the duration of the unidentified state with the threshold of the duration of the unidentified state. If the duration of the unidentified state is less than or equal to the threshold of the duration of the unidentified state, perform an off-duty behavior analysis on the first tracking person and obtain the first off-duty behavior analysis sub-result; otherwise, determine that the corresponding first tracking person is off-duty, reduce the misjudgment of off-duty behavior, and make the off-duty analysis process more reasonable.

[0111] In one embodiment, if the duration of the unrecognized state is less than or equal to a preset threshold of the duration of the unrecognized state, an off-duty behavior analysis is performed on the first tracked person, and a first off-duty behavior analysis sub-result is obtained, including:

[0112] Attempt to detect a substitute guardian in the second working area; wherein the substitute guardian is a staff member who temporarily replaces the first tracking staff member to work in the corresponding second working area;

[0113] If the substitute guardian attempts to obtain the information successfully, the first off-duty behavior analysis sub-result of the first tracking person in the second work area is output as not off-duty;

[0114] If the substitute guardian fails in his attempt to obtain it, the corresponding second working area will be used as the third working area;

[0115] Try to obtain the temporary leave request of the first tracking person corresponding to the third work area; the temporary leave request is: a request to actively report temporary leave, such as: temporarily leave for 5 minutes, please system hosting;

[0116] If the temporary leave request is successfully obtained, the first off-duty behavior analysis sub-result of the first tracked person in the third work area is output as not off-duty;

[0117] If the temporary leave request fails to be obtained, the corresponding third work area is used as the fourth work area, and at the same time, the first tracking person corresponding to the fourth work area is used as the second tracking person;

[0118] Obtaining a shooting blind area of ​​the fourth working area; wherein the shooting blind area is: a portion of the fourth working area that is not captured after all captured videos of the fourth working area are merged;

[0119] According to the shooting blind area, determine the blind area information acquisition strategy; wherein the blind area information acquisition strategy is: to determine whether there is a person in the blind area to obtain relevant information;

[0120] Obtain blind spot information according to a blind spot information acquisition strategy; wherein the blind spot information is: relevant information for determining whether there is a person in the blind spot;

[0121] Determine whether the second tracking person is in the corresponding fourth working area according to the blind area information;

[0122] If the second tracked person is in the corresponding fourth work area, outputting the first off-duty behavior analysis sub-result of the second tracked person in the corresponding fourth work area as not off-duty;

[0123] If the second tracked person is not in the corresponding fourth work area, the first off-duty behavior analysis sub-result of the second tracked person in the corresponding fourth work area is output as off-duty.

[0124] The working principle and beneficial effects of the above technical solution are:

[0125] When the duration of the unrecognized state is less than or equal to a preset threshold of the duration of the unrecognized state, the present invention attempts to obtain a substitute guardian in the second working area. If there is a corresponding substitute, it is determined that there is no absence from the post. If it cannot be determined that there is a substitute guardian, a temporary leave request of the first tracking person is determined. If the request is successfully obtained, it is also determined that the person is not off-duty. In addition, considering the shooting blind spot of the fourth working area, a blind spot information acquisition strategy is determined and corresponding blind spot information is acquired. According to the blind spot information, it is determined whether the second tracking person is in the corresponding fourth working area. According to the different blind spot information judgment results, corresponding output is performed, and the off-duty judgment is performed taking into account the working status of various situations, which is more comprehensive.

[0126] In one embodiment, attempting to detect a replacement guardian in the second work area includes:

[0127] Detecting the first tracking person in the second working area other than the first tracking person corresponding to the second working area as the third tracking person;

[0128] Obtaining the previous communication information between the third tracking person and the first tracking person corresponding to the second working area; wherein the previous communication information is: the interaction information between the first tracking person and the third tracking person before it is detected that the first tracking person corresponding to the second working area is not in the corresponding area, including interaction actions, interaction dialogues, etc.;

[0129] If the previous communication information meets the substitute communication conditions, and there is a fourth tracker in the second work area that the third tracker is responsible for, the third tracker is determined to be the substitute guardian in the corresponding second work area; wherein the second work area that the third tracker is responsible for is: the work area that the third tracker was originally responsible for; the fourth tracker is: the first tracker who is responsible for the second work area that the third tracker is responsible for, except the third tracker;

[0130] Among them, the conditions for substitute exchange include:

[0131] The preceding communication action includes at least one first confirmation action by a third tracking person; wherein the first confirmation action is: an approval action by the third tracking person;

[0132] and,

[0133] After the first confirmation action, at least one second confirmation action of the fourth tracking person can be obtained; the time interval between the first confirmation action and the second confirmation action is less than or equal to a preset action time interval threshold; wherein the second confirmation action is: the approval action of the fourth tracking person; the action time interval threshold is manually preset, for example: 5 seconds;

[0134] and,

[0135] The pre-order communication semantics includes the preset time representation semantics and substitute matter representation semantics. Among them, the time representation semantics is: time description semantics, such as: "Please substitute for me for 5 minutes"; the substitute matter representation semantics is: "Please substitute for me", "Substitute for me for a shift".

[0136] The working principle and beneficial effects of the above technical solution are:

[0137] The present invention obtains the previous communication information of the third tracking person and the first tracking person corresponding to the second working area, introduces the substitute communication condition to perform the substitute determination;

[0138] Specifically, the conditions include:

[0139] The preceding communication action includes at least one first confirmation action by a third tracker, indicating that the third tracker recognizes the substitute behavior; after the first confirmation action, at least one second confirmation action by a fourth tracker can be obtained; the action time interval between the first confirmation action and the second confirmation action is less than or equal to the preset action time interval threshold; indicating that the fourth tracker is aware of the substitute behavior of the third tracker who works together; the preceding communication semantics includes the preset time representation semantics and substitute matter representation semantics, indicating that the communication matter is indeed a substitute matter, and contains time description information, which is in compliance with the specification.

[0140] The present invention obtains the previous communication information between the third tracking person and the first tracking person corresponding to the second work area, introduces a plurality of substitute communication conditions to perform substitute determination, and improves the standardization of substitute determination.

[0141] In one embodiment, according to the shooting blind area, determining the blind area information acquisition strategy includes:

[0142] Obtaining scanning point cloud data of the fourth working area; wherein the scanning point cloud data is: a result obtained by scanning the fourth working area in advance using a three-dimensional scanning device;

[0143] Constructing a scanning point cloud model of the fourth working area according to the scanning point cloud data; wherein the scanning point cloud model is: a three-dimensional building model of the fourth working area;

[0144] Based on the current target camera parameters, a virtual camera is set in the scanned point cloud model; wherein the target camera parameters are: the camera parameters of the target camera, used to simulate the settings of the virtual camera for shooting, including parameters such as focal length, aperture, shutter speed, etc.; the virtual camera is: a software tool that simulates a real camera in 3D modeling;

[0145] Determine a shooting blind area in the scanned point cloud model; wherein the shooting blind area is a mapping result of a corresponding area of ​​the fourth working area that cannot be captured by the target camera in the scanned point cloud model, and perform three-dimensional reconstruction based on the fourth working area captured by the target camera and then compare and obtain it with the scanned point cloud model;

[0146] According to the virtual camera, simulated shooting of the shooting blind area is performed to obtain a simulated shooting result; wherein the simulated shooting result is: a simulated shooting image obtained by the virtual camera simulating shooting of the shooting blind area;

[0147] If the simulated shooting result is successful, determine the simulated shooting parameters of the virtual camera; wherein the simulated shooting parameters of the virtual camera are: the setting parameters of the virtual camera when the simulated shooting result is successful;

[0148] Shooting in the blind area is performed according to the virtual camera simulation shooting parameters and the target camera parameters; wherein, when shooting in the blind area, the target camera parameters are set according to the parameter conversion relationship between the virtual camera simulation shooting parameters and the target camera parameters before shooting;

[0149] If the simulated shooting result is a shooting failure, a verification reminder is sent to the second tracking person in the corresponding fourth working area, wherein the verification reminder is a reminder to remind the second tracking person to go out of the blind area for verification.

[0150] The working principle and beneficial effects of the above technical solution are:

[0151] The present invention obtains scanning point cloud data of the fourth working area and constructs a scanning point cloud model, and sets a virtual camera in the scanning point cloud model based on the current target camera parameters to simulate shooting of the blind area and obtain a simulated shooting result; when the simulated shooting result is successful, it means that the target camera is able to shoot the blind area, but the current shooting strategy setting is not suitable, and the target camera parameters can be set according to the virtual camera simulation shooting parameters; if the simulated shooting result fails, it means that the current hardware configuration cannot achieve shooting of the blind area, and the second tracking personnel is reminded to go out of the blind area for verification, which is more appropriate.

[0152] In one embodiment, simulating shooting of a shooting blind area with a virtual camera to obtain a simulation shooting result includes:

[0153] Based on the preset shooting feature extraction template, the shooting features are extracted, and the shooting features include: the current simulation parameters of the virtual camera and the blind area information features of the shooting blind area; wherein the preset shooting feature extraction template is used to extract the shooting feature template according to the shooting task; the current simulation parameters include: camera setting position parameters, shooting parameters, etc.; the blind area information features are: the feature representation of the blind area information, including: the blind area range, the relative position relationship between the blind area and the camera, etc.;

[0154] Based on the historical shooting records of the target camera, a shooting model is trained; wherein the historical shooting records are: a record of the shooting process of the target camera in history; the shooting model is: an AI model that learns the historical shooting records and adaptively determines the best shooting setting parameters according to the shooting target;

[0155] The simulation shooting results are obtained according to the current simulation parameters, blind spot information characteristics and shooting model.

[0156] The working principle and beneficial effects of the above technical solution are:

[0157] The present invention introduces a shooting feature extraction template to characterize the shooting task and obtain shooting features. In addition, the historical shooting records of the target camera are introduced to train the shooting model, and the simulated shooting results are obtained according to the current simulation parameters, blind spot information features and the shooting model, thereby improving the efficiency of obtaining the simulated shooting results, and further improving the speed of determining the blind spot information acquisition strategy.

[0158] The embodiment of the present invention provides a railway trestle video intelligent analysis application system based on oil depot automatic control signal triggering, such as Figure 2 As shown, including:

[0159] The oil depot automatic control signal identification subsystem 1 is used to try to obtain the oil depot automatic control signal during the oil depot railway oil delivery operation;

[0160] The railway trestle video shooting subsystem 2 is used to trigger multiple target cameras on the trestle platform to shoot the trestle platform and obtain the railway trestle video if the acquisition attempt is successful;

[0161] Off-post behavior analysis subsystem 3, used to analyze the off-post behavior of the guardian based on the railway trestle video based on the intelligent video analysis technology, and obtain the off-post behavior analysis results;

[0162] The analysis result warning subsystem 4 is used to output visualization results and issue warnings based on the off-duty behavior analysis results.

[0163] In one embodiment, the oil depot automatic control signal identification subsystem includes:

[0164] A communication link establishment module is used to establish a communication link with the communication node of the host computer in the central control room;

[0165] The oil depot automatic control signal acquisition module is used to obtain the oil depot automatic control signal based on the communication link.

[0166] In one embodiment, the railway trestle video shooting subsystem includes:

[0167] A control template acquisition module is used to obtain the control template of the target camera if the acquisition attempt is successful;

[0168] The railway trestle video acquisition module is used to control the target camera to shoot the trestle platform based on the control template and acquire the railway trestle video.

[0169] 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. The railway trestle video intelligent analysis application method based on oil depot automatic control signal triggering is characterized by: include: Try to obtain the oil depot automatic control signal during the oil depot railway oil delivery operation; If the acquisition attempt is successful, multiple target cameras on the trestle platform are triggered to shoot the trestle platform to obtain the railway trestle video; Based on intelligent video analysis technology, the off-duty behavior of supervisors is analyzed based on the railway trestle video to obtain the off-duty behavior analysis results; Based on the analysis results of off-duty behavior, visual results are output and warnings are issued.

2. The railway trestle video intelligent analysis application method based on oil depot automatic control signal triggering as claimed in claim 1 is characterized in that: Try to obtain the oil depot automatic control signal during the oil depot railway oil delivery operation, including: Establish a communication link with the communication node of the host computer in the central control room; Based on the communication link, the oil depot automatic control signal is obtained.

3. The railway trestle video intelligent analysis application method based on oil depot automatic control signal triggering according to claim 1 is characterized in that: If the acquisition attempt is successful, multiple target cameras on the trestle platform are triggered to shoot the trestle platform and obtain the railway trestle video, including: If the acquisition attempt is successful, obtain the control template of the target camera; Based on the control template, the target camera is controlled to shoot the trestle platform to obtain the railway trestle video.

4. The railway trestle video intelligent analysis application method based on oil depot automatic control signal triggering as claimed in claim 1 is characterized in that: Based on intelligent video analysis technology, the off-duty behavior of supervisors is analyzed based on the railway trestle video, and the off-duty behavior analysis results are obtained, including: Based on YOLO v8, target detection and target tracking are performed on the supervisors in the railway trestle video to obtain detection and tracking results; Conduct off-duty behavior analysis based on the detection and tracking results to obtain the off-duty behavior analysis results.

5. The railway trestle video intelligent analysis application method based on oil depot automatic control signal triggering as claimed in claim 4 is characterized in that: Perform off-duty behavior analysis based on the detection and tracking results, and obtain off-duty behavior analysis results, including: According to the detection and tracking results, detecting whether there is a corresponding supervisor in each first working area of ​​the trestle platform; If there is no corresponding guardian in the first working area, the corresponding first working area is used as the second working area, and the corresponding non-existent guardian is used as the first tracking person; Obtaining the duration of the unidentified state of the first tracked person in the corresponding second working area; If the duration of the unrecognized state is less than or equal to the preset threshold of the duration of the unrecognized state, an off-duty behavior analysis is performed on the first tracked person to obtain a first off-duty behavior analysis sub-result; If the duration of the unrecognized state is longer than the preset threshold of the duration of the unrecognized state, the second off-duty behavior analysis sub-result of the first tracked person is output; The first off-duty behavior analysis sub-result and the second off-duty behavior analysis sub-result are integrated to obtain the off-duty behavior analysis result.

6. The railway trestle video intelligent analysis application method based on oil depot automatic control signal triggering as claimed in claim 5 is characterized in that: If the duration of the unrecognized state is less than or equal to the preset threshold of the duration of the unrecognized state, an off-duty behavior analysis is performed on the first tracked person, and a first off-duty behavior analysis sub-result is obtained, including: Attempt to detect a backup guardian in a secondary work area; If the substitute guardian attempts to obtain the information successfully, the first off-duty behavior analysis sub-result of the first tracking person in the second work area is output as not off-duty; If the substitute guardian fails in his attempt to obtain it, the corresponding second working area will be used as the third working area; Try to obtain the temporary leave request of the first tracking person corresponding to the third work area; If the temporary leave request is successfully obtained, the first off-duty behavior analysis sub-result of the first tracked person in the third work area is output as not off-duty; If the temporary leave request fails to be obtained, the corresponding third work area is used as the fourth work area, and at the same time, the first tracking person corresponding to the fourth work area is used as the second tracking person; Obtain the shooting blind area of ​​the fourth working area; Determine the strategy for acquiring blind spot information based on the blind spots in shooting; Obtain blind spot information according to a blind spot information acquisition strategy; Determine whether the second tracking person is in the corresponding fourth working area according to the blind area information; If the second tracked person is in the corresponding fourth work area, outputting the first off-duty behavior analysis sub-result of the second tracked person in the corresponding fourth work area as not off-duty; If the second tracked person is not in the corresponding fourth work area, the first off-duty behavior analysis sub-result of the second tracked person in the corresponding fourth work area is output as off-duty.

7. The railway trestle video intelligent analysis application method based on oil depot automatic control signal triggering as claimed in claim 6 is characterized in that: Attempt to detect backup supervisors in the secondary work area, including: Detecting the first tracking person in the second working area other than the first tracking person corresponding to the second working area as the third tracking person; Obtaining the previous communication information between the third tracking person and the first tracking person corresponding to the second working area; If the previous communication information meets the substitute communication conditions, and there is a fourth tracker in the second work area that the third tracker is responsible for, the third tracker is determined to be the substitute guardian in the corresponding second work area; Among them, the conditions for substitute exchange include: The preceding communication action includes at least one first confirmation action by a third tracking person; and, After the first confirmation action, at least a second confirmation action of a fourth tracking person can be obtained; the time interval between the first confirmation action and the second confirmation action is less than or equal to a preset time interval threshold; and, Pre-order communication semantics includes the semantics of pre-set time representation and the semantics of substitute event representation.

8. Railway trestle video intelligent analysis application system based on oil depot automatic control signal triggering, characterized in that: include: The oil depot automatic control signal identification subsystem is used to try to obtain the oil depot automatic control signal during the oil depot railway oil dispatching operation; The railway trestle video shooting subsystem is used to trigger multiple target cameras on the trestle platform to shoot the trestle platform and obtain the railway trestle video if the acquisition attempt is successful; The off-post behavior analysis subsystem is used to analyze the off-post behavior of the supervisor based on the railway trestle video based on the intelligent video analysis technology and obtain the off-post behavior analysis results; The analysis result warning subsystem is used to output visual results and issue warnings based on the off-duty behavior analysis results.

9. The railway trestle video intelligent analysis application system based on oil depot automatic control signal triggering as claimed in claim 8, characterized in that: The oil depot automatic control signal identification subsystem includes: A communication link establishment module is used to establish a communication link with the communication node of the host computer in the central control room; The oil depot automatic control signal acquisition module is used to obtain the oil depot automatic control signal based on the communication link.

10. The railway trestle video intelligent analysis application system based on oil depot automatic control signal triggering according to claim 8, characterized in that: Railway trestle video shooting subsystem, including: A control template acquisition module is used to obtain the control template of the target camera if the acquisition attempt is successful; The railway trestle video acquisition module is used to control the target camera to shoot the trestle platform based on the control template and acquire the railway trestle video.

Images