A method and system for detecting oil theft incidents in parking lots using real-time video streaming

Abstract

The present invention discloses a method and system for detecting oil theft incidents in parking lots using real-time video streams, belonging to the field of vehicle safety detection technology. The method comprises the following steps: acquiring real-time video, obtaining vehicle status information based on vehicle information in the real-time video, and determining whether the vehicle is in a stopped state; acquiring the distance between a first vehicle and a second vehicle, both of which are in a stopped state, and determining that a suspected oil theft incident exists when the distance between the first vehicle and the second vehicle is less than a preset threshold; timing the status of the first vehicle and the second vehicle in which the suspected oil theft incident exists, and determining that a confirmed oil theft incident exists when the timing exceeds a first preset time length; and issuing a warning message. The present invention proposes a new method for determining vehicle driving status based on the motion characteristics of vehicles in video streams. The method extracts the vehicle's position changes based on a target tracking algorithm, and reflects the vehicle's current motion state based on the continuous position change results.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle safety detection, and in particular relates to a method and system for detecting oil theft incidents by vehicles in parking lots using real-time video streaming. Background Art

[0002] Truck fuel tanks are generally large, especially for super-heavy and heavy trucks, which have a fuel tank capacity of at least 500 liters and can reach up to 1,000 liters.

[0003] Due to the large number of trucks, their wide distribution, the large amount of fuel in their tanks, and the presence of some criminals, truck fuel theft occurs frequently, causing heavy losses to truck drivers. In order to avoid stopping for rest, which provides opportunities for oil thieves, some drivers choose to stop for rest less frequently or even not at all, which indirectly aggravates fatigue driving and increases the risk of road accidents. Summary of the Invention

[0004] The present invention provides a method and system for detecting oil theft incidents by vehicles in parking lots using real-time video streaming, which can effectively detect oil theft incidents by vehicles and avoid the risk of oil theft.

[0005] The present invention is achieved through the following technical solutions:

[0006] On the one hand, the present invention provides a method for detecting oil theft incidents in vehicles in a parking lot using real-time video streaming, comprising the following steps: acquiring real-time video, acquiring vehicle status information based on vehicle information in the real-time video, and determining whether the vehicle is in a stopped state; acquiring the distance between a first vehicle and a second vehicle, both of which are in a stopped state, and determining that a suspected oil theft incident exists when the distance between the first vehicle and the second vehicle is less than a preset threshold; timing the status of the first vehicle and the second vehicle in which the suspected oil theft incident occurs, and determining that a confirmed oil theft incident exists when the timing exceeds a first preset time length; and issuing a warning message.

[0007] In some embodiments, timing the status of the first vehicle and the second vehicle in which the oil theft incident is suspected comprises the following steps: when the first vehicle and the second vehicle change from a state in which the oil theft incident is suspected to be present to a normal state and then change again to a state in which the oil theft incident is suspected within a second preset time period, resetting the current timing status.

[0008] In some embodiments, after determining that an oil theft event has occurred, the following steps are further included: presetting an initial lifetime for determining the oil theft event; judging each frame of image after determining that an oil theft event has occurred to determine whether an oil theft event has occurred; decrementing the initial lifetime by 1 for each frame of image in which no oil theft event occurs; and determining that the oil theft event has ended when the initial lifetime reaches 0.

[0009] In some embodiments, the process of decrementing the initial lifespan by 1 each time there is no oil theft incident in a frame of image further includes the following steps: in consecutive preset frame images, the first vehicle and the second vehicle are restored to a state where an oil theft incident is confirmed, and the current lifespan is reset to the initial lifespan.

[0010] In some embodiments, obtaining real-time video, obtaining vehicle status information based on vehicle information in the real-time video, and determining whether the vehicle is in a stopped state include the following steps: initializing the video stream; obtaining the vehicle's trajectory line; and obtaining vehicle status information based on the vehicle's trajectory line.

[0011] In some embodiments, obtaining the trajectory of a vehicle includes the following steps: obtaining the detected position of the jth vehicle in the i-th frame of the video:

[0012]

[0013] in, To detect the x coordinate of the upper left corner of the position box; To detect the y coordinate of the upper left corner of the position box; Detect the x coordinate of the lower right corner of the location box; Detect the y coordinate of the lower right corner of the position frame; get the center position of the jth vehicle in the i-th frame image. The center position is:

[0014]

[0015] The trajectory line of the j-th vehicle is obtained, where the line connecting the center coordinates of the detection position boxes of the j-th vehicle's consecutive frames is the trajectory line of the vehicle.

[0016] In some embodiments, obtaining vehicle status information based on the vehicle's trajectory includes the following steps: calculating the distance between the center position of the j-th vehicle detection position frame in the i+1-th frame image and the j-th vehicle detection position frame corresponding to the i-th frame image in the i+1-th frame image; if the distance is less than a certain threshold θ for a continuous period of time t, the vehicle is determined to be in a stopped state; otherwise, it is in a moving state.

[0017] In some embodiments, obtaining vehicle status information includes: obtaining a large vehicle and all vehicles that are less than a certain distance from the large vehicle; obtaining the distance between a first vehicle and a second vehicle that are both in a stopped state, where the first vehicle is a large vehicle.

[0018] In some embodiments, the distance between a first vehicle and a second vehicle, both of which are in a stopped state, is obtained. When the distance between the first vehicle and the second vehicle is less than a preset threshold, in the process of determining that a suspected oil theft incident exists, determining the distance between the first vehicle and the second vehicle includes the following steps: forming a preset range with the first vehicle as the center; determining the center position of the second vehicle; drawing a ray toward the center of the first vehicle with the center position of the second vehicle as the origin; when the intersection of the ray with the edge of the preset range is an odd number, determining that the distance between the first vehicle and the second vehicle is less than the preset threshold; when the intersection of the ray with the edge of the preset range is an even number, determining that the distance between the first vehicle and the second vehicle is not less than the preset threshold.

[0019] On the other hand, the present application provides a real-time video stream parking lot vehicle oil theft incident detection system, including a video acquisition module, a processor and a memory, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, it implements the real-time video stream parking lot vehicle oil theft incident detection method of any one of the above embodiments.

[0020] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0021] (1) The method for detecting oil theft incidents in parking lots using real-time video streams provided by the present invention proposes a new method for determining the vehicle's driving state based on the motion characteristics of vehicles in video streams. The method extracts the vehicle's position changes based on a target tracking algorithm and reflects the vehicle's current motion state based on the continuous position change results.

[0022] (2) The method for detecting oil theft incidents in parking lots using real-time video streams provided by the present invention is based on target detection and tracking algorithms, combines the position coordinates and motion status of the vehicle, and adopts an unsupervised, real-time dynamic identification method for oil theft incidents based on the characteristics of oil theft incidents in video streams. The algorithm is efficient and concise, and has higher operability.

[0023] (3) The method for detecting oil theft incidents in parking lots using real-time video streams provided by the present invention utilizes a method that integrates multiple mechanisms to determine oil theft incidents. The algorithm is relatively accurate and effective, and can determine the end state of the oil theft incident. It has strong robustness and higher accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings in the embodiments will be briefly introduced below. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1A flowchart of a method for detecting oil theft incidents in parking lots using real-time video streams provided by an embodiment of the present invention;

[0026] Figure 2 Schematic diagram of the process of obtaining vehicle status information provided by different embodiments of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0028] In the description of the present invention, it should be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc. used to indicate the orientation or position relationship are based on the orientation or position relationship shown in the accompanying drawings, or are the orientation or position relationship in which the product of the invention is usually placed when in use. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, they should not be understood as limiting the present invention.

[0029] Furthermore, the use of terms such as "horizontal" and "vertical" in the description of the present invention does not necessarily imply that the component must be absolutely horizontal or suspended, but rather that it can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal than "vertical" and does not mean that the structure must be completely horizontal, but rather that it can be slightly tilted.

[0030] It should also be noted that, in the description of the present invention, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal connections between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention in specific contexts.

[0031] The terms "comprise," "comprising," and "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or modules is not limited to the listed steps or modules, but may optionally include steps or modules not listed, or may optionally include other steps or modules inherent to the process, method, product, or apparatus.

[0032] In one aspect, an embodiment of the present application provides a method for detecting oil theft incidents in parking lots using real-time video streams, comprising the following steps:

[0033] S10: Acquire real-time video, and based on the vehicle information in the real-time video, obtain vehicle status information to determine whether the vehicle is stopped. In S10, information is obtained using real-time video to determine whether the vehicles in the current parking lot are stopped. The vehicle being determined can be a single vehicle, multiple vehicles, or even all vehicles in the parking lot. By determining whether the vehicle is stopped, further determination is made based on the vehicle being stopped to determine whether an oil theft incident has occurred.

[0034] Specifically, S10 may include the following steps:

[0035] S101. Video stream initialization. In S101, the regions within each frame of the real-time video can be divided under any task, and the location coordinates of the selected regions for that task can be obtained. This facilitates oil theft detection for vehicles in different regions corresponding to different tasks. In a specific example, the selected regions can be part of or all of the corresponding frame images in the real-time video. Furthermore, video stream initialization can also initialize and set parameters required for vehicle oil theft detection, such as, for example, distance thresholds and time thresholds between vehicles that will be used later. Another example is region attributes, which facilitates data reading and storage.

[0036] S102, obtaining vehicle trajectories. In S102, based on task requirements, such as all vehicles, some vehicles, or specified vehicles in a selected area, the trajectories of these vehicles are obtained.

[0037] Exemplarily, obtaining the vehicle's trajectory may specifically include the following steps:

[0038] S1021. Obtain a detected position of the jth vehicle in the i-th frame image of the real-time video. The detected position may be a position box including the j-th vehicle. The i-th frame image and the j-th vehicle are both general terms and may refer to any frame image and any vehicle in the image.

[0039] For example, the position frame is described as a rectangular frame. A rectangular frame centered on the jth vehicle in the i-th frame image is obtained. The four coordinate points of the rectangular frame are mainly detected to obtain the detection position information of the jth vehicle in the i-th frame image:

[0040]

[0041] in, To detect the x coordinate of the upper left corner of the position box; To detect the y coordinate of the upper left corner of the position box; Detect the x coordinate of the lower right corner of the location box; The y-coordinate of the lower right corner of the detection position frame. Connect the detection positions of the jth vehicle in the consecutive frames, including the i-th frame, to form the vehicle's trajectory. The number of consecutive frames can be set as needed.

[0042] For example, when connecting the detection positions, the center position of the obtained detection position information of the j-th vehicle can be used as a reference. The center position of the j-th vehicle in the i-th frame image is:

[0043]

[0044] The center positions of the jth vehicle in consecutive frames, including the i-th frame, are then connected to form the vehicle's trajectory. By contrast, the trajectory obtained by connecting the center positions more intuitively and accurately reflects the vehicle's trajectory. In other examples, a point on the edge of the vehicle's location box, such as a corner point of a rectangular box, can be used as a reference to form the trajectory.

[0045] S103. Obtain vehicle status information based on the vehicle's trajectory. For example, still taking the jth vehicle as an example, after obtaining the jth vehicle's trajectory, calculate the distance between the vehicle's detected position in the (i+1)th frame image and the vehicle's detected position in the i-th frame image. If the calculated corresponding distances between two adjacent frames of real-time video are both less than a certain threshold value θ within a continuous period of t, then the vehicle is determined to be in a stopped state. If, within a continuous period of t, the calculated corresponding distances between two adjacent frames of real-time video are greater than or equal to the threshold value θ, then the vehicle is determined to be in a moving state.

[0046] S20: Further determine whether there has been an oil theft incident for the parked vehicles. Specifically, the distance between a first and second parked vehicle in the selected area is obtained. The first and second vehicles can be any two vehicles in the selected area. By determining whether there has been an oil theft incident between any two vehicles, it is determined whether there has been an oil theft incident between all vehicles in the selected area. When the distance between the first and second vehicles is less than a preset threshold, a suspected oil theft incident is determined. The preset threshold for this distance is pre-set, and the specific threshold value can be set based on the current frequency of oil theft incidents in the parking lot and can be modified at any time.

[0047] S20 may specifically include the following steps:

[0048] S201: Obtaining the distance between a first vehicle and a second vehicle, both of which are in a stopped state. Specifically, S201 includes the following steps:

[0049] S2011: A preset range is formed with the first vehicle as the center. In S2011, the first vehicle is used as the center, primarily with the center position of the first vehicle as a reference. The center position of the first vehicle can be obtained in the manner described in step S1021. The preset range can be a regular shape, such as a circle or rectangle, formed with the first vehicle as the center.

[0050] S2012: Determine the center position of the second vehicle. In S2012, the center position of the second vehicle can also refer to the method in step S1021.

[0051] S2013: Draw a ray toward the center of the first vehicle with the center position of the second vehicle as the origin.

[0052] S2014: Determine the intersection of the ray obtained in step S2013 with the edge of the preset range formed in step S2011. Based on the number of intersections between the ray and the edge of the preset range, determine whether the distance between the second vehicle and the first vehicle is less than a preset threshold. Specifically, if the number of intersections between the ray and the edge of the preset range is an odd number, the second vehicle is within the preset range formed by the first vehicle as the center, and the distance between the first and second vehicles is determined to be less than the preset threshold. If the number of intersections between the ray and the edge of the preset range is an even number, the distance between the first and second vehicles is determined to be not less than the preset threshold.

[0053] S202: When the distance between the first vehicle and the second vehicle is less than a preset threshold, it is determined that a suspected oil theft event occurs.

[0054] In some embodiments, in S10, when acquiring vehicle status information, the information is primarily collected for large vehicles and other vehicles within a certain distance of the large vehicle. In S20, the first vehicle determined to have been the victim of oil theft is a large vehicle, and the large vehicle is used as the first vehicle to determine whether an oil theft incident has occurred. Because large vehicles are more susceptible to oil theft incidents, targeting large vehicles for a specific group and detecting oil theft incidents can effectively increase the system's computing power, thereby improving processing efficiency.

[0055] S30: When there is a suspected oil theft incident between the first vehicle and the second vehicle, further judgment is performed.

[0056] Exemplarily, S30 may include the following steps:

[0057] S301: Further determination can be made using a countdown mechanism, with a first preset duration set. If a suspected oil theft incident persists between the first vehicle and the second vehicle in consecutive frames of imagery during the first preset duration, a confirmed oil theft incident is determined. This example configuration allows for tracking suspected oil theft incidents and further determination, avoiding misdiagnosis and false alarms and improving system robustness.

[0058] In some embodiments, S30 may further include the following steps:

[0059] S302: A penalty mechanism is further employed. When the first and second vehicles are determined to be in a suspected oil theft incident, a countdown mechanism is employed for further determination. If the first and second vehicles transition from a suspected oil theft incident state to a normal state, i.e., within a first preset time period, there is at least one frame of consecutive frames in which the first and second vehicles do not maintain a suspected oil theft incident state, then the suspected oil theft incident cannot be determined as a confirmed oil theft incident and continued observation is required. At this point, the timing state in S301 is reset. That is, after the first and second vehicles no longer maintain a suspected oil theft incident state, if the normal state transitions again from a suspected oil theft incident state within a second preset time period, starting from the current frame image, the timing is restarted based on the first preset time period. With this configuration, if, in consecutive frame images, the first and second vehicles transition from a suspected oil theft incident to a normal state and then again become suspected oil theft incidents, restarting the judgment at step S10 would significantly increase the system's computational process, resulting in reduced computational performance and a slower final result. Therefore, with the configuration of this embodiment, the penalty mechanism is employed to directly continue the judgment at step S301, effectively improving the system's processing efficiency and enabling faster processing results.

[0060] S40: After obtaining data confirming the oil theft incident, issuing an early warning message.

[0061] In some embodiments, after step S30, the following steps are further included:

[0062] S50: Determine whether the oil theft incident has ended. For example, S50 specifically includes:

[0063] S501. Pre-set an initial lifespan for determining oil theft events. In S501, the pre-set initial lifespan can be specifically set based on the response of the detection station, i.e., the response speed for determining oil theft events, and the frequency of oil theft events in the parking lot to be detected.

[0064] S502: After the confirmed oil theft event is determined, each frame of image is judged to determine whether the confirmed oil theft event exists. After the confirmed oil theft event exists, each frame of image in which the relative positions of the first vehicle and the second vehicle change, that is, when the state of the confirmed oil theft event changes from the state of the confirmed oil theft event to the normal state in the image frame, the confirmed oil theft event does not end immediately. Instead, a process of determining the end of the event is performed. Specifically, each time such a frame of image exists, the initial lifespan is decremented by 1 until the lifespan reaches 0, at which point the oil theft event is determined to be ended, and a signal for ending the event is issued.

[0065] S50 can be performed simultaneously with S40. Specifically, after the oil theft incident is confirmed, S40 and S50 are performed simultaneously. While issuing the warning information, the end of the oil theft incident is determined. Before the end of the oil theft incident is determined, the warning information can be continuously issued. After the end of the oil theft incident is determined, the end of the incident information is issued, and then the warning information stops being issued, and the current oil theft incident ends.

[0066] In some embodiments, S50 further includes the following steps:

[0067] S503: Establish a resurrection mechanism. During S502, the first and second vehicles transition from a normal state to a state indicating an oil theft incident. Starting from the frame at which the transition to an oil theft incident is established, consecutive frames of images, each of which is in the state indicating an oil theft incident, are used. The specific number of consecutive frames can be pre-set. At this point, the current lifespan is reset to the initial lifespan, and S502 is repeated to determine the end of the oil theft incident. Establishing a resurrection mechanism improves the accuracy of oil theft incident detection and enhances system robustness.

[0068] By setting up the countdown mechanism, penalty mechanism, lifespan mechanism and resurrection mechanism, the occurrence and end of the oil theft incident can be comprehensively determined, making the system's judgment more accurate, effectively reducing false alarms and improving the robustness of the system.

[0069] On the other hand, an embodiment of the present application provides a real-time video stream parking lot vehicle oil theft incident detection system, including a video acquisition module, a processor and a memory, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, it implements the real-time video stream parking lot vehicle oil theft incident detection method of any one of the above embodiments.

[0070] An embodiment of the present application further provides a computer storage medium having a computer program stored thereon, which is loaded by a processor to execute the method for detecting oil theft incidents by vehicles in parking lots using real-time video streams according to any one of the above embodiments.

[0071] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any simple modification or equivalent change made to the above embodiment based on the technical essence of the present invention shall fall within the scope of protection of the present invention.

Claims

1. A method for detecting oil theft incidents in parking lots using real-time video streams, characterized in that: The following steps are involved: Acquire real-time video, and based on vehicle information in the real-time video, acquire vehicle status information to determine whether the vehicle is in a stopped state; Obtaining a distance between a first vehicle and a second vehicle, both of which are in a stopped state, and determining that a suspected oil theft incident occurs when the distance between the first vehicle and the second vehicle is less than a preset threshold; timing the status of the first vehicle and the second vehicle suspected of having the oil theft incident, and determining that the oil theft incident has been confirmed when the timing exceeds a first preset time period; Issue early warning information; The timing of the status of the first vehicle and the second vehicle in which the suspected oil theft incident occurs comprises the following steps: When the first vehicle and the second vehicle change from a state in which a suspected oil theft incident occurs to a normal state and change again to a state in which a suspected oil theft incident occurs within a second preset time period, the current timing state is reset; After determining that there is an oil theft incident, the following steps are also included: An initial life span for determining oil theft events is pre-set; After determining that an oil theft event exists, each frame of the image is judged to determine whether the oil theft event exists; Every time there is no oil theft event in a frame of image, the initial lifespan is reduced by 1; When the initial lifespan is reduced to 0, it is determined that the oil theft event is over; The process of decrementing the initial lifespan by 1 for each frame of image in which no oil theft event occurs further includes the following steps: In the continuous frame images, the first vehicle and the second vehicle are restored to a state where the oil theft event is confirmed, and the current lifespan is reset to the initial lifespan.

2. The method for detecting oil theft incidents in parking lots using real-time video streams according to claim 1 is characterized in that: The acquiring of real-time video, acquiring vehicle status information based on vehicle information in the real-time video, and determining whether the vehicle is in a stopped state comprises the following steps: Video stream initialization; Get the vehicle's trajectory; Vehicle status information is obtained based on the trajectory of the vehicle.

3. The method for detecting oil theft incidents in parking lots using real-time video streams according to claim 2 is characterized in that: The method of obtaining the vehicle's trajectory comprises the following steps: Get video i Frame image j Detection location of each vehicle: in, To detect the upper left corner of the position box x coordinate; To detect the upper left corner of the position box y coordinate; Detection location box lower right corner x coordinate; Detection position box lower right corner y coordinate; Get the i Frame image j The center position of the vehicle is: Get the j The trajectory of the vehicle, j The line connecting the center coordinates of the detection position frames of consecutive vehicle frames is the trajectory line of the vehicle.

4. The method for detecting oil theft incidents in parking lots using real-time video streams according to claim 3 is characterized in that: The obtaining of vehicle status information based on the vehicle trajectory includes the following steps: Calculate the i +1 frame image j The center position of the vehicle detection position frame is i The frame image is described in i +1 frame image corresponding to j The distance between the vehicle detection position frame, if the distance is continuous t If the time is less than the threshold θ, the vehicle is determined to be in a stopped state, otherwise it is in a moving state.

5. The method for detecting oil theft incidents in parking lots using real-time video streams according to claim 1 is characterized in that: The vehicle status information is obtained by: obtaining a large vehicle and all vehicles whose distance from the large vehicle is less than a calibrated distance; In the acquiring of the distance between a first vehicle and a second vehicle, both of which are in a stopped state, the first vehicle is a large vehicle.

6. The method for detecting oil theft incidents by vehicles in parking lots using real-time video stream according to claim 1 or 4, characterized in that: The process of obtaining the distance between the first vehicle and the second vehicle, both of which are in a stopped state, and determining that a suspected oil theft incident has occurred when the distance between the first vehicle and the second vehicle is less than a preset threshold, comprises the following steps: A preset range is formed with the first vehicle as the center; a center position of the second vehicle is determined; a ray is drawn with the center position of the second vehicle as the origin toward the center of the first vehicle; When the number of intersections between the ray and the edge of the preset range is odd, it is determined that the distance between the first vehicle and the second vehicle is less than a preset threshold; When the number of intersections between the ray and the edge of the preset range is even, it is determined that the distance between the first vehicle and the second vehicle is not less than a preset threshold.

7. A real-time video streaming parking lot vehicle oil theft incident detection system, characterized in that: The method comprises a video acquisition module, a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the method for detecting oil theft incidents of vehicles in parking lots using real-time video streams as claimed in any one of claims 1 to 6 is implemented.

Images