A passenger car overload detection method, device and electronic equipment

Abstract

The present invention discloses a method, device and electronic equipment for detecting overload of a passenger car, and specifically relates to the field of detection technology. The method comprises obtaining head information in a car; performing scale mapping on the head information to obtain mapped head information; judging whether the mapped head information is within a calibration area; and if the mapped head information is within the calibration area, prompting that the passenger car is overloaded. By obtaining the head information in the car, mapping the acquired head information, and then comparing and judging the mapped head information with the calibration area, it is determined whether there is head information within the calibration area, and judging whether the passenger car is overloaded according to the head information. The head mapping solves the problem that the position of the passenger's head or face in the image cannot accurately reflect the specific position of the passenger in the car due to the limited height of the car, thereby realizing accurate detection of overload of the passenger car.

Description

Technical Field

[0001] The present invention relates to the field of detection technology, and in particular to a bus overload detection method, device and electronic equipment. Background Art

[0002] With the development of the economy and society and the increase in the floating population, the problem of bus overloading has become increasingly prominent. Bus overloading is an important cause of frequent traffic accidents. The main manifestation of bus overloading is that some passengers stay in the aisle area of ​​the carriage for a long time. Therefore, the use of smart devices to detect passengers in the aisle area of ​​the carriage has become an effective way to determine whether the bus is overloaded.

[0003] The existing passenger car overload detection method is to install intelligent monitoring equipment in the car, analyze and process the video data collected by the camera, so as to determine whether there are passengers in the aisle, so as to determine whether the vehicle is overloaded. For example, by manually calibrating the aisle area, a detector is used to detect the passengers in the calibrated area. The detector can give a position boundary box of the passenger's head, face or even the whole body, and judge whether the passenger is in the aisle area according to the position relationship between the boundary box and the calibrated area. Although this method can detect passengers in the aisle area of ​​the car to a certain extent, due to the limited height in the car, the camera cannot be installed in the top view position, but can only be installed in the head of the car. Its viewing angle has a certain angle, resulting in the position of the passenger's head or face in the image cannot accurately reflect the specific position of the passenger in the car, thereby causing the detector to be unable to accurately detect the overload of the passenger car when performing overload detection. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a bus overload detection method, device and electronic equipment to solve the problem that the prior art cannot accurately detect bus overload.

[0005] According to the first aspect, an embodiment of the present invention provides a method for detecting bus overload, comprising: obtaining head information in a car; performing scale mapping on the head information to obtain mapped head information; determining whether the mapped head information is within a calibrated area; if the mapped head information is within the calibrated area, prompting that the bus is overloaded.

[0006] The bus overload detection method provided by the embodiment of the present invention obtains head information in the car, maps the obtained head information, and then compares and judges the mapped head information with the calibration area, so as to determine whether there is head information in the calibration area, and judge whether the bus is overloaded according to the head information. The head mapping solves the problem that the position of the passenger's head or face in the image cannot accurately reflect the specific position of the passenger in the car due to the limited height of the car, thereby realizing accurate detection of bus overload.

[0007] In combination with the first aspect, in the first implementation of the first aspect, if the mapping header information is in the calibrated area, it is prompted that the bus is overcrowded, and it also includes: based on a preset time length, statistics are kept of the time length information of the mapping header information staying in the calibrated area; it is determined whether the time length information of the mapping header information staying in the calibrated area is greater than the threshold information; if the time length information is greater than the threshold information, it is prompted that the bus is overcrowded; if the time length information is less than or equal to the threshold information, the bus is kept under overload detection.

[0008] The bus overload detection method provided by the embodiment of the present invention presets a time length, and uses the preset time length to statistically map the time length information that the head information stays in the calibration area, and then determines whether the time length information exceeds the threshold information, thereby ensuring in one step that the bus overload situation can be accurately detected. Setting the preset time length can also reduce the detector's misjudgment of the bus overload situation, thereby improving the detection accuracy of the bus overload detection method.

[0009] In combination with the first aspect or the first embodiment of the first aspect, in the second embodiment of the first aspect, the obtaining of the head information in the car includes: obtaining a video image frame in the car; performing preprocessing based on the video image frame to obtain a preprocessed image frame; sending the preprocessed image frame to a head information detection model for detection to obtain the head information in the car, wherein the head information in the car includes: a head boundary box, coordinate information of the detection area, and confidence.

[0010] In combination with the second embodiment of the first aspect, in the third embodiment of the first aspect, the preprocessing is performed based on the video image frame to obtain the preprocessed image frame, including: dividing the video image frame in the vehicle compartment into detection areas to obtain at least 2 detection areas; intercepting and / or scaling based on the detection areas to obtain the preprocessed image frame.

[0011] The bus overload detection method provided by the embodiment of the present invention further improves the detection accuracy and the model detection speed by preprocessing the video image frames.

[0012] In combination with the third implementation of the first aspect, in the fourth implementation of the first aspect, before the head information is scale-mapped to obtain the mapped head information, it also includes: filtering the head bounding box according to the confidence level, and / or screening to obtain a filtered head bounding box; determining the coordinate information of the filtered head bounding box based on the coordinate information of the detection area; and restoring the filtered head bounding box coordinate information to the video image frame to obtain the head information.

[0013] The bus overload detection method provided by the embodiment of the present invention ensures the accuracy of the detection result and improves the detection efficiency by eliminating the boundary boxes that do not meet the detection requirements before performing scale mapping.

[0014] In combination with the first aspect or the first implementation of the first aspect, in the fifth implementation of the first aspect, the scale mapping of the head information to obtain the mapped head information includes: obtaining the side length value of the calibrated area in the video image frame, the actual width value corresponding to the calibrated area, and the ordinate value of the calibrated area in the video image frame; based on the side length value of the calibrated area in the video image frame, the actual width value corresponding to the calibrated area, and the ordinate value of the calibrated area in the video image frame, using a mapping equation to calculate the mapping distance of the head information; and moving the head information according to the mapping distance of the head information to obtain the mapped head information. .

[0015] The bus overload detection method provided by the embodiment of the present invention uses a mapping equation to map the head information, so as to accurately determine the location of the passengers and improve the accuracy of detection.

[0016] In combination with the first aspect or the first embodiment of the first aspect, in the sixth embodiment of the first aspect, the determination of whether the mapped header information is within the calibrated area includes: obtaining two fixed points of the middle line of the head boundary frame in the mapped header information; calculating the degree of overlap between the two fixed points and the calibrated area; and determining whether the mapped header information is within the calibrated area based on the degree of overlap.

[0017] The bus overload detection method provided by the embodiment of the present invention determines whether the head information is within the calibrated area by calculating the overlap between the two fixed points of the middle line of the head boundary frame and the calibrated area, thereby being able to quickly determine whether the bus is overloaded and further improve the detection efficiency.

[0018] According to the second aspect, an embodiment of the present invention provides a bus overload detection device, including: an acquisition module, used to acquire head information in the car; a mapping module, used to scale-map the head information to obtain mapped head information; a judgment module, used to judge whether the mapped head information is within a calibrated area; and a prompt module, used to prompt that the bus is overloaded if the mapped head information is within the calibrated area.

[0019] The bus overload detection device provided in the embodiment of the present invention obtains the header information through the acquisition module, and then maps the header information using the mapping module. After that, the mapped header information is sent to the judgment module, and its calibration area is compared with the header information to determine whether the header information is within the calibration area, thereby realizing accurate detection of bus overload.

[0020] According to the third aspect, an embodiment of the present invention provides an electronic device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the bus overload detection method described in the first aspect or any one of the embodiments of the first aspect by executing the computer instructions.

[0021] According to the fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable the computer to execute the passenger car overload detection method described in the first aspect or any one of the embodiments of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the present invention in any way. In the accompanying drawings:

[0023] Figure 1 It is a flowchart of a bus overload detection method provided by an embodiment of the present invention;

[0024] Figure 2 is a flowchart of step S10 of an optional bus overload detection method provided by an embodiment of the present invention;

[0025] Figure 3 It is a flowchart of step S11 of an optional bus overload detection method provided by an embodiment of the present invention;

[0026] Figure 4 is a schematic diagram of a mapping in an optional bus overload detection method provided by an embodiment of the present invention;

[0027] Figure 5 is a flowchart of step S12 of an optional bus overload detection method provided in an embodiment of the present invention;

[0028] Figure 6 It is a flowchart of steps S141 to S143 of an optional bus overload detection method provided by an embodiment of the present invention;

[0029] Figure 7 is a flowchart of step S13 of an optional bus overload detection method provided in an embodiment of the present invention;

[0030] Figure 8 It is a structural block diagram of a passenger car overload detection device provided by an embodiment of the present invention;

[0031] Fig. 9 It is a structural block diagram of an electronic device provided by an embodiment of the present invention.

[0032] Reference numerals

[0033] 10 - acquisition module; 11 - mapping module; 12 - judgment module; 13 - prompt module; 20 - processor; 21 - memory. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution 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 part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0035] In addition, it should be noted that a bus overload detection method provided by an embodiment of the present invention can be applied in the scenario of bus compartment overload, and the method can be implemented by an overload detection device or equipment, wherein the device for implementing the bus overload detection method may include: an acquisition module, a mapping module, a judgment module and a prompt module; wherein the acquisition module may also include a video acquisition submodule and a head information detection module; the video acquisition submodule can be used for video data in the bus compartment, wherein the video data may be real-time video data; the video acquisition submodule is connected to the head information detection module, and is used to output the head information in the compartment; the mapping module includes a bounding box mapping submodule, which is used to scale the head information; the judgment module is used to acquire the mapped head information and extract the pre-stored calibration area from the memory for comparison to determine whether the head information belongs to an overloaded head information; in addition, the prompt module may also include an alarm submodule and an upload submodule, the alarm submodule is used to perform a bus overcrowding alarm, and the upload submodule is used to perform data storage backtracking. Specifically, an embodiment of the present invention provides a bus overload detection method, such as Figure 1 As shown, the bus overload detection method comprises:

[0036] S10, obtaining head information in the car.

[0037] In this embodiment, the head information in the carriage can be obtained by acquiring the video data in the carriage and sending the data into the head detection model. The construction of the head detection model can be to train the DCNN head detection model through manually annotated data, and embed the trained model into the smart device. When the video acquisition device processes the video frame, the model will be automatically loaded, and each frame of the video image will be detected. The head detection model can give the bounding box coordinates of all passenger heads in the video frame, as well as the confidence score that the bounding box is the head, and filter and screen the bounding box by threshold and confidence, so as to reduce false detection and improve detection accuracy.

[0038] S11, scale mapping the header information to obtain mapped header information.

[0039] In this embodiment, due to the use of non-top view cameras, the position of the head information of a person in the image will be quite different under different human postures. For example, when a passenger is standing, the head position is close to the top of the image, and when squatting, the head position is close to the bottom of the image, so the head bounding box cannot accurately reflect the position of the passenger in the carriage. Therefore, in this embodiment, the detected head bounding box is mapped, and the mapped position can be on the ground of the carriage aisle, thereby improving the accuracy of the carriage aisle passenger detection.

[0040] S12, determining whether the mapping header information is within the calibration area.

[0041] In this embodiment, the mapped head information is a head bounding box, and the calibrated area is a preset range area, which is the aisle area of ​​the carriage in this embodiment. Whether the bus is overloaded is determined by judging whether the head bounding box is in the aisle of the carriage. Therefore, by calculating the overlap between the mapped frame of the head information and the calibrated area to judge whether the passenger is in the aisle, the passenger's location can be judged more accurately.

[0042] S13, if the mapped head information is within the calibrated area, it is prompted that the bus is overloaded.

[0043] In this embodiment, when the bus is overcrowded, a voice reminder can be given to the driver through the prompt module, and the reminder can be saved as an alarm attachment and uploaded to the supervision cloud platform.

[0044] The bus overload detection method provided by the embodiment of the present invention obtains head information in the car, maps the obtained head information, and then compares and judges the mapped head information with the calibration area, so as to determine whether there is head information in the calibration area, and judge whether the bus is overloaded according to the head information. The head mapping solves the problem that the position of the passenger's head or face in the image cannot accurately reflect the specific position of the passenger in the car due to the limited height of the car, thereby realizing accurate detection of bus overload.

[0045] Optional, such as Figure 2 As shown, step S10 may also be:

[0046] S101, acquiring video image frames in the vehicle compartment.

[0047] In this embodiment, the video image frames in the vehicle compartment may be video data collected by video acquisition devices installed at four corners of the vehicle compartment, and the video data is split to obtain video image frames.

[0048] S102, performing preprocessing based on the video image frame to obtain a preprocessed image frame.

[0049] In this embodiment, the preprocessing is to divide the video image frame in the vehicle into detection areas to obtain at least two detection areas; based on the detection areas, the video image frame is captured and / or scaled to obtain a preprocessed image frame. Optionally, the head information in the vehicle may include: a head boundary box, coordinate information of the detection area, and a confidence level.

[0050] S103, sending the pre-processed image frame to the head information detection model for detection to obtain the head information in the car.

[0051] In this embodiment, the head information detection model can be a pre-built model, wherein the model can be a detection model built by a convolutional neural network. Since the input resolution of the detection model used is relatively small, if the original image is used as the input file, the original image will become blurred after being reduced, and more useful feature information will be lost, which will result in a low detection rate of the detection model, and may even fail to detect the passengers in the back row of the car. In order to improve the detection accuracy of the model, this embodiment divides the detection area of ​​the original image, and divides the image into two upper and lower detection areas, wherein the aspect ratio of the upper detection area is kept consistent with the model input aspect ratio. Before the detection model performs detection, the two detection areas need to be cropped and scaled respectively, and input into the model respectively, and the head detection is performed on the two detection areas. After that, the detection results need to be restored to the original image according to the coordinate information of the detection area to ensure the accuracy of the data. By dividing the detection area, the detection capability of the rear passengers is increased, and the recall rate and accuracy of the head detection in the aisle of the car are greatly improved.

[0052] Optional, such as Figure 3 As shown, step S11 may also be:

[0053] S111, obtaining a side length value of the calibrated area in the video image frame, an actual width value corresponding to the calibrated area, and a vertical coordinate value of the calibrated area in the video image frame.

[0054] In this embodiment, the aisle in the car can be manually or automatically calibrated using calibration software or manual calibration to obtain the side length value of the aisle in the car, the actual width value corresponding to the calibrated area and the vertical coordinate value of the calibrated area in the video image frame.

[0055] S112, calculating the mapping distance of the head information by using a mapping equation based on the side length value of the calibrated area in the video image frame, the actual width value corresponding to the calibrated area, and the vertical coordinate value of the calibrated area in the video image frame.

[0056] In this embodiment, before obtaining the mapping distance, it is also necessary to ensure that the acquisition camera is installed horizontally and is located at the top of the middle of the front of the car. The camera is facing the aisle and has a small downward angle. This ensures the accuracy of the detection results. In addition, the car aisle area manually calibrated on the video image is also required, for example: Figure 4 As shown, quadrilateral ABCD, when the camera installation conditions are met, the quadrilateral ABCD in this area is an isosceles trapezoid or approximates an isosceles trapezoid, that is, AB is parallel to DC, and AD is equal to BC.

[0057] Then, a mapping equation is established according to the obtained calibration area. The specific expression of the mapping equation is:

[0058] y′=(w1 / w)*k+b,

[0059] Among them, w1 is the average side length of the head bounding box in the image, w is the average width of the actual human head, k and b are the mapping distances, y′ is the ordinate of the centroid of the mapping bounding box, and the abscissa of the centroid of the mapping bounding box remains unchanged. According to the scale mapping relationship of the calibrated isosceles trapezoidal area in the image, the equation group is established as follows:

[0060]

[0061] Among them, w AB is the length of AB in the image, w CD is the length of CD in the image, w o is the actual width of the aisle, y1 is the ordinate of AB in the image, y2 is the ordinate of CD in the image, and the mapping distances k and b are obtained using the equations:

[0062]

[0063] After mapping the centroid of the head bounding box through the above mapping equation, the ordinate y′ of the new centroid is obtained, while the abscissa remains unchanged, and the length and width of the mapped bounding box remain unchanged. The mapped bounding box can accurately represent the position of the passenger on the floor of the carriage. By comparing the overlap between the mapped bounding box and the calibrated area, it can accurately determine whether the passenger is in the aisle of the carriage.

[0064] S113: Move the header information according to the mapping distance of the header information to obtain the mapped header information.

[0065] In this embodiment, after the mapping distance is obtained, the mapping header information can be obtained by moving in a translation manner.

[0066] Optional, such as Figure 5 As shown, step S12 may include:

[0067] S121, obtaining two fixed points on the midline of the head boundary frame in the mapped head information.

[0068] In this embodiment, two fixed points of the center line of the head boundary frame need to be obtained by manual marking, or they can be marked by using marking software.

[0069] S122, calculating the overlap between the two fixed points and the calibration area.

[0070] S123: Determine whether the mapped header information is within the calibration area based on the degree of overlap.

[0071] In this embodiment, the overlap between the mapping frame and the calibration area is calculated, and the threshold can be used to control the sensitivity of the alarm. The specific method is: first find the two fixed points on the middle line of the mapping frame, which can be recorded as EF for ease of understanding, and use the pnpoly algorithm to determine whether the fixed points E and F are in the calibration area. Specifically: when E and F are both in the calibration area, the overlap is 1; when E and F are not in the calibration area and are located on the same side of the calibration area, the overlap is 0; when point E is in the calibration area and point F is outside, calculate the coordinates of the intersection O of EF and the calibration area, and the overlap is r=EO / EF; when point E is outside the area and point F is inside the area, the overlap is r=OF / EF. By setting the threshold δ, when r>δ, it can be considered that the passenger is in the aisle of the carriage. Optionally, the sensitivity of the alarm can be adjusted by adjusting the threshold.

[0072] In this embodiment, whether there are passengers in the aisle of the carriage is determined by calculating the overlap, thereby improving the detection accuracy.

[0073] Optional, such as Figure 6 As shown, before step S11, the following steps may also be included:

[0074] S141, filtering and / or screening the head bounding box according to the confidence level to obtain a screened head bounding box.

[0075] In this embodiment, the confidence level is used to filter and / or screen the head bounding box, thereby reducing error parameters, reducing the probability of false detection, and improving detection accuracy.

[0076] S142 determines the coordinate information of the filtered head boundary box based on the coordinate information of the detection area.

[0077] In this embodiment, the coordinate information is used to ensure that the filtered head boundary box can be restored to the video frame of the original image, thereby ensuring the accuracy of detection.

[0078] S143, restore the filtered head boundary box coordinate information to the video image frame to obtain the head information.

[0079] The embodiment of the present invention provides a method for detecting overloading of a passenger car, comprising: Figure 1-5 In addition to the steps shown, step S13 also includes: Figure 7 As shown:

[0080] S131, based on the preset duration statistics, map the duration information of the header information staying in the marked area.

[0081] In this embodiment, the time length is manually preset, and the total time or image frame number of the mapped head frame staying in the calibration area within the preset time length is counted. The time length information of the mapped head information staying in the calibration area includes: the total time the mapped head information stays, for example: 20 minutes, and the total number of image frames of the mapped head information staying in the calibration area within the preset time length, for example: there are 100 image frames in 20 minutes, and the image frames of the mapped head information account for 70 image frames.

[0082] S132, determining whether the duration information of the mapping header information staying in the calibration area is greater than the threshold information.

[0083] S133: If the duration information is greater than the threshold information, it is indicated that the bus is overcrowded.

[0084] In this embodiment, the reminder of the bus overcrowding may be provided by a voice reminder to the driver through a voice device, or the overcrowding information or video may be uploaded to the supervision center.

[0085] S134: If the duration information is less than or equal to the threshold information, the overload detection of the bus is continued.

[0086] In this embodiment, maintaining overload detection on the passenger car may be reacquiring head information in the car.

[0087] An embodiment of the present invention provides a method for detecting overloading of a passenger car, the method comprising:

[0088] First, a head detection model based on DCNN is established. The detection device can automatically load the model and process the aisle video collected by the camera in real time to detect the passenger head boundary box.

[0089] Then, the scale-based mapping method is used to map the head bounding box to the ground where the passenger is located, and the positional relationship between the mapped bounding box and the calibration area is determined to determine whether the passenger is in the aisle of the carriage; finally, the total time when there are passengers in the aisle within a time period is counted. If it is greater than the threshold, a voice alarm is issued through the buzzer to remind the driver, and the alarm attachment is uploaded to the network cloud platform. The specific process is as follows:

[0090] First, the original video frame in the car is collected, and its image frame is divided into two upper and lower detection areas; the two detection areas are respectively cropped and scaled, and then the detection areas are input into the pre-trained head detection model for detection. The head detection model can give the coordinate information and confidence of the head bounding box based on the detection area; then the confidence threshold is used to filter and select the head bounding box; then, according to the coordinates of each detection area, the coordinates of the filtered bounding box are restored to the original image to ensure the integrity of the data; then, the mapping distance in the mapping equation is calculated through the manually calibrated coordinates of the aisle area of ​​the car; then, based on The mapping distance head bounding box is mapped to the ground of the lane aisle; then the overlap between all mapped head bounding boxes and the calibration area is calculated one by one. If there is an overlap greater than the threshold, it is considered that there is a passenger in the aisle of the carriage in the video frame. If the preset time is greater than the threshold at this time, the flag variable t is set to 1, otherwise t is set to 0; finally, a fixed-length sequence is established. After processing a frame of image data, the y value is stored in the sequence; then the sum of all y values ​​in the sequence is calculated. If it is greater than the threshold, an alarm is given that there is a passenger in the aisle of the carriage, and the driver is reminded by voice. The video segment is saved as an alarm attachment and uploaded to the supervision cloud platform.

[0091] The bus overload detection method provided in this embodiment first divides the original image into two detection areas to detect the head information in the car, thereby improving the recall rate and accuracy of head detection;

[0092] Secondly, this detection method manually calibrates the carriage aisle area and adopts a scale mapping method to map the head boundary box to the ground of the lane aisle, thereby improving the detection accuracy of passengers in the carriage aisle. After that, the method provided in this embodiment detects through a sequence of preset time length or preset specified length. When the preset time threshold is reached, if there are still passengers in the carriage aisle, the overloading behavior of the vehicle is monitored and alarmed, thereby ensuring the safety of vehicle driving.

[0093] The embodiment of the present invention provides a bus overload detection device, such as Figure 8 As shown, the bus overload detection device comprises:

[0094] The acquisition module 10 is used to acquire the head information in the vehicle compartment. For details, please refer to the relevant description of step S10 of the above method embodiment.

[0095] The mapping module 11 is used to perform scale mapping on the header information to obtain mapped header information. For details, refer to the relevant description of step S11 in the above method embodiment.

[0096] The judgment module 12 is used to judge whether the mapping header information is within the calibration area. For details, please refer to the relevant description of step S12 of the above method embodiment.

[0097] The prompt module 13 is used to prompt that the bus is overcrowded if the mapped header information is within the calibration area. For details, refer to the relevant description of step S13 of the above method embodiment.

[0098] The bus overload detection device provided in this embodiment obtains header information through an acquisition module, and then maps the header information using a mapping module. The mapped header information is then sent to a judgment module, and its calibration area is compared with the header information to determine whether the header information is within the calibration area, thereby achieving accurate detection of bus overload.

[0099] In addition, an embodiment of the present invention further provides an electronic device, such as Fig. 9 As shown, the electronic device may include a processor 20 and a memory 21, wherein the processor 20 and the memory 21 may be connected via a bus or other means. Fig. 9 The example of connecting through bus is taken in the following.

[0100] The processor 20 may be a central processing unit (CPU). The processor 20 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above chips.

[0101] The memory 21 is a non-transitory computer-readable storage medium that can be used to store non-transitory software programs, non-transitory computer executable programs and modules, such as program instructions / modules corresponding to the method for shielding buttons of a vehicle-mounted display device in an embodiment of the present invention (for example, Figure 8 The processor 20 executes various functional applications and data processing of the processor by running the non-transient software programs, instructions and modules stored in the memory 21, that is, the bus overload detection method in the above method embodiment is implemented.

[0102] The memory 21 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required by at least one function; the data storage area may store data created by the processor 20, etc. In addition, the memory 21 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 21 may optionally include a memory remotely arranged relative to the processor 20, and these remote memories may be connected to the processor 20 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0103] The one or more modules are stored in the memory 21, and when executed by the processor 20, the following is performed: Figure 1-7 The passenger car overload detection method in the illustrated embodiment.

[0104] For details of the above electronic equipment, please refer to Figures 1 to 8 The corresponding related descriptions and effects in the illustrated embodiments can be understood and will not be repeated here.

[0105] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD) or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above-mentioned types of memory.

[0106] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for detecting overloading of a passenger car, characterized in that: include: Get the head information in the car; Performing scale mapping on the header information to obtain mapped header information; Determining whether the mapping header information is within a calibration area; If the mapped header information is within the calibrated area, a prompt is given that the bus is overloaded; The step of obtaining the head information in the carriage includes: Acquire video image frames in the vehicle compartment; Performing preprocessing based on the video image frame to obtain a preprocessed image frame; Sending the preprocessed image frame into a head information detection model for detection to obtain the head information in the carriage, wherein the head information in the carriage includes: a head boundary box, coordinate information of a detection area, and a confidence level; The step of performing scale mapping on the header information to obtain mapped header information includes: Obtaining a side length value of the calibrated area in the video image frame, an actual width value corresponding to the calibrated area, and a ordinate value of the calibrated area in the video image frame; Calculating the mapping distance of the head information by using a mapping equation based on the side length value of the calibration area in the video image frame, the actual width value corresponding to the calibration area, and the ordinate value of the calibration area in the video image frame; The header information is moved according to a mapping distance of the header information to obtain the mapped header information.

2. The method according to claim 1, characterized in that If the mapped header information is within the calibrated area, prompting that the bus is overloaded also includes: Counting the time information of the mapping header information staying in the marked area based on a preset time length; Determine whether the time length information of the mapping header information staying in the calibration area is greater than the threshold information; If the duration information is greater than the threshold information, a prompt is given that the bus is overloaded; If the duration information is less than or equal to the threshold information, the overload detection of the bus is continued.

3. The method according to any one of claim 1, characterized in that: The preprocessing based on the video image frame to obtain the preprocessed image frame includes: Dividing the video image frame in the vehicle compartment into detection areas to obtain at least two detection areas; The pre-processed image frame is obtained by performing interception and / or scaling based on the detection area.

4. The method according to claim 3, characterized in that: Before the header information is scale-mapped to obtain the mapped header information, the method further includes: Filtering and / or screening the head bounding box according to the confidence level to obtain a screened head bounding box; Based on the coordinate information of the detection area, determining the coordinate information of the filtered head bounding box; The filtered head boundary box coordinate information is restored to the video image frame to obtain the head information.

5. The method according to claim 1, characterized in that The determining whether the mapping header information is within the calibration area includes: Obtain two fixed points on the midline of the head bounding box in the mapped head information; Calculating the degree of overlap between the head bounding box and the calibration area according to the positional relationship between the two fixed points and the calibration area; It is determined whether the mapping header information is within a calibration area according to the overlap degree.

6. A passenger car overload detection device, characterized in that: include: An acquisition module is used to acquire head information in the carriage; A mapping module, used for performing scale mapping on the header information to obtain mapped header information; A judging module, used to judge whether the mapping header information is within a calibration area; A prompting module, used for prompting that the bus is overloaded if the mapped header information is within the calibration area; The step of obtaining the head information in the carriage includes: Acquire video image frames in the vehicle compartment; Performing preprocessing based on the video image frame to obtain a preprocessed image frame; Sending the preprocessed image frame into a head information detection model for detection to obtain the head information in the carriage, wherein the head information in the carriage includes: a head boundary box, coordinate information of a detection area, and a confidence level; The step of performing scale mapping on the header information to obtain mapped header information includes: Obtaining a side length value of the calibrated area in the video image frame, an actual width value corresponding to the calibrated area, and a ordinate value of the calibrated area in the video image frame; Calculating the mapping distance of the head information by using a mapping equation based on the side length value of the calibration area in the video image frame, the actual width value corresponding to the calibration area, and the ordinate value of the calibration area in the video image frame; The header information is moved according to a mapping distance of the header information to obtain the mapped header information.

7. An electronic device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the bus overload detection method according to any one of claims 1 to 5 by executing the computer instructions.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the passenger car overload detection method according to any one of claims 1-5.

Images