Automatic vehicle alignment method and autonomous navigation batch vehicle inspection method for vehicle bottom security inspection robot

Abstract

The invention discloses an automatic vehicle alignment method and an autonomous navigation batch vehicle inspection method for a vehicle bottom security inspection robot, which are characterized in that a to-be-detected vehicle is accurately identified through combination of a front camera real-time video stream and a target detection algorithm, a target vehicle is selected through an app terminal, the position of the target vehicle in an image is automatically calculated and converted into a laser radar coordinate, and the vehicle bottom security inspection robot is automatically aligned with the vehicle bottom security inspection robot. Accurate scanning of the vehicle contour is realized. And the robot speculates the vehicle position according to a laser radar scanning result, automatically aligns and starts a vehicle bottom scanning task. At a user side, a parking lot model can be constructed through an app. The selectable detection modes comprise a single train mode, a longitudinal train mode, a transverse train mode and a multi-row self-defined mode. The vehicle bottom security inspection robot can execute corresponding autonomous navigation batch vehicle inspection operation according to different modes selected by the user so as to meet diversified security inspection requirements.

Description

Technical Field

[0001] The invention relates to safety detection technology, and in particular to an automatic vehicle alignment method of an under-vehicle security inspection robot and an autonomous navigation batch inspection method for vehicles. Background Art

[0002] The undercarriage security inspection robot is an automated device specially used for safety inspection of the undercarriage of vehicles. It is usually equipped with high-definition cameras, sensors, lighting equipment, etc., and can perform comprehensive and detailed scanning and inspection of the undercarriage without direct human contact with the vehicle. The undercarriage security inspection robot replaces manual inspection of the undercarriage, can quickly complete the undercarriage inspection, improve the efficiency of security inspection, and reduce the waiting time of vehicles. Through advanced image recognition and analysis technology, it can accurately detect abnormal conditions under the vehicle, such as hidden objects, modified parts, etc. It avoids direct contact between security inspectors and the undercarriage of vehicles that may be dangerous, ensuring the safety of security inspectors. It can operate autonomously according to preset procedures and paths, reducing human errors.

[0003] However, the current undercar inspection robots may have some shortcomings, such as the inability to autonomously realize vehicle alignment detection and return, which requires manual operation; after inspecting a vehicle, it needs to be manually moved to another adjacent vehicle, and the autonomous batch inspection capability is limited. The robot's navigation strategy is not strong or flexible enough, and it cannot autonomously move to adjacent vehicles to realize batch inspection. The inability to automatically detect multiple rows of vehicles and other issues have limited its wide application and performance to a certain extent.

[0004] The existence of the above problems leads to many deficiencies in the use of existing under-vehicle security inspection robots: Low security inspection efficiency: After each vehicle inspection, the robot needs to be moved manually, which takes a lot of time and reduces the overall security inspection speed, especially in places with large vehicle traffic, which is prone to congestion.

[0005] Increased labor costs: Relying on manual labor to move the robot increases manpower input and increases operating costs.

[0006] Increased safety risks: Since it is not possible to automatically detect multiple rows of vehicles, the underbody inspection of some vehicles may be missed, which poses a safety risk. In addition, since it is not possible to automatically align the center of the vehicle to be tested, the chassis image acquisition may be incomplete.

[0007] Poor work continuity: The parking lot cannot be mapped in advance, which makes the robot's work lack planning and continuity, affecting the overall security inspection effect.

[0008] Decreased service quality: An unsmooth and inconvenient security check process will affect user experience and reduce the quality and satisfaction of related services.

[0009] Weak emergency response capabilities: In emergency situations where a large number of vehicles need to be quickly checked, such as major events or emergencies, the task cannot be completed efficiently. Summary of the invention

[0010] In view of the deficiencies in the prior art, the present invention aims to provide an automatic vehicle alignment method for an under-vehicle security inspection robot and an autonomous navigation batch vehicle inspection method, which can automatically align with the center of a vehicle to be tested and realize batch scanning inspection of vehicle chassis.

[0011] The present invention is achieved through the following technical solutions: A method for automatically aligning an under-vehicle security inspection robot with a vehicle comprises the following steps: Step 1: After the robot starts, it receives the real-time video stream from the front camera and performs image processing internally. It runs the target detection algorithm on the obtained data to identify all the vehicles in the image. Each identified vehicle is marked with a rectangular box (x1, y1, x2, y2); Step 2: The video stream data of the front camera is transmitted to the app via a wireless network; Step 3: After the user clicks on the vehicle to be tested on the screen on the app display interface, the click coordinate c1 will be sent to the robot. Based on the user's click coordinate c1, the robot program determines which rectangular box is the vehicle selected by the operator, and then calculates the lateral position range of the vehicle in the image, expressed as c2 = (xl, xr), which is the horizontal coordinates of the left and right edges of the vehicle on the image; Step 4: Convert the c2 value in the image coordinate system to the actual physical coordinate system of the laser radar. After completing the coordinate conversion, calculate the corresponding angle interval (a1, a2) that the laser radar should scan. Step 5: The laser radar scans according to the starting angle a1 and the ending angle a2, searches for the left and right wheels of the vehicle, and obtains their coordinate positions c3 and c4; the position of the vehicle to be tested is inferred from the positions c3 and c4, and after obtaining the position of the vehicle to be tested, automatic alignment is performed to start the underbody scanning task.

[0012] Further, the process of automatic alignment described in step 5 is as follows: Step 1: When the robot receives the double-click command, it enters the automatic vehicle alignment mode; Step 2: The robot scans the front wheel track of the current vehicle through the laser radar, and selects the center axis of the two-wheel connection line to find the target point 0.5m away from the robot side; Step 3: Adjust the extension line of the robot's head to align with the target point and move to the target point; Step 4: Adjust the robot angle on the spot so that its body is completely aligned with the mid-perpendicular line of the target path.

[0013] A method for autonomously navigating batch inspection of vehicles by an under-vehicle security inspection robot, which builds a parking lot model and a detection model through a user-side app. The detection modes include longitudinal train mode, transverse vehicle mode, and multi-row custom mode. Train mode: S1.1. The user inputs the total number of vehicles in the longitudinal train, and operates the app to move the undercarriage inspection robot to the front position of the first vehicle in the longitudinal train. When the front camera identifies the first vehicle to be inspected through the target recognition algorithm, the laser radar is used to scan the front wheel track of the first vehicle, and the center axis of the double-wheel connection line is selected to find the first target point 0.5m away from the robot side; S1.2, the under-vehicle security inspection robot reaches the first target point, so that the wheel tracks on both sides of the laser radar are the same, and the license plate information is recognized by the front camera. The robot moves forward and passes under the first vehicle. When the distance measurement component on the top of the vehicle detects that a vehicle has been inspected according to the distance change, the complete under-vehicle image captured by the linear array camera component is obtained, and a stop command is issued to the robot to stop; S1.3, the total number of vehicles in the longitudinal train is reduced by one, and the laser radar is again aimed at the next vehicle in a similar way to identify the license plate information. Regardless of whether the license plate is found, the undercarriage inspection robot scans the wheelbase in front through the laser radar and calibrates the vehicle again. After the calibration is completed, the next vehicle is inspected until all vehicles are completely inspected and the undercarriage images of each vehicle are obtained; Horizontal car mode: S2.1. Use the app to move the undercarriage inspection robot to the front of the first car in the longitudinal train. The robot's front camera identifies the vehicle to be inspected through the target recognition algorithm. S2.2. Scan the front wheel track with the laser radar and find the first target point 3m away from the center axis of the two-wheel connection line to the robot side. S2.3. Autonomously navigate to the first target point, rotate the robot in place so that the wheelbases of both sides of the laser radar are the same. At this time, the robot aims at the first vehicle and recognizes the license plate. The robot moves forward and passes under the first vehicle. When the top distance measurement component determines that the robot has completely passed under the vehicle, the system obtains a complete picture of the vehicle bottom and then retreats back along the original route. S2.4, the robot returns to the posture aimed at the first vehicle and determines whether all vehicle inspections have been completed. If completed, it jumps to step S2.6; if not, it executes step S2.5; S2.5, turn 90 degrees to the side of the next vehicle, and then drive forward a fixed distance D, where the fixed distance D is calculated by the formula Calculated, where W represents the width of the parking space, n represents the number of empty parking spaces. After reaching the target point, immediately rotate 90 degrees in the opposite direction and loop through steps S2.1 to S2.4; S2.6, end the task, output and store all vehicle license plates and complete chassis image information; Multi-row custom mode: When executing the multi-row mode, select one of the longitudinal train mode or the transverse vehicle mode to execute the single-row mode. After executing the single-row, rotate 180 degrees and complete the scanning of the opposite row of vehicles in reverse, end the task, and output and store all vehicle license plates and complete chassis picture information.

[0014] Furthermore, in the horizontal parking mode, the user first inputs the number of horizontal parking spaces and parking space width information on the app, marks the empty parking spaces, and then generates a horizontal parking batch inspection model.

[0015] The present invention has the following beneficial effects: (1) By combining the real-time video stream of the front camera with the target detection algorithm, the vehicle to be tested can be accurately identified and the target vehicle can be selected through the app. The position of the target vehicle in the image is automatically calculated and converted into LiDAR coordinates to accurately scan the vehicle outline. The robot infers the vehicle position based on the LiDAR scanning results and automatically aligns to start the underbody scanning task, which greatly improves the automation, accuracy and efficiency of underbody security inspection, reduces manual operations, and improves work convenience and safety.

[0016] (2) It can realize batch scanning and inspection of vehicle chassis, avoiding the situation where the system needs to be manually moved to another adjacent vehicle after inspecting one vehicle each time; (3) In the application scenario of vehicle chassis inspection, most people do not have the ability to build SLAM maps of parking lots in advance. The present invention only requires users to input a small number of parameters to achieve batch inspection of vehicles in mainstream standard parking lots, thereby improving work efficiency.

[0017] (4) In this technical solution, the alignment algorithm of a single vehicle is corrected according to the wheelbase of the front two wheels, which can accurately align the vehicle. Every time a new vehicle is scanned sideways, the algorithm will be used again for correction, avoiding cumulative errors. This ensures the success rate of the system's autonomous batch vehicle inspection tasks.

[0018] (5) The present invention is scalable and can realize the inspection of multiple rows of vehicles through simple combination tasks on the application side. For most standard parking lots, it can semi-autonomously complete the scanning of all vehicles in the entire parking lot without the need for map construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a schematic diagram of the structure of the under-vehicle security inspection robot; Figure 2 It is a schematic diagram of the longitudinal train mode; Figure 3 It is the flow chart of the longitudinal train mode detection algorithm; Figure 4 This is a schematic diagram of the horizontal vehicle mode; Figure 5 This is the flow chart of the horizontal vehicle mode detection algorithm; Figure 6 This is a schematic diagram of the multi-row vehicle mode.

[0020] In the figure: 1- linear array camera assembly, 2- top ranging assembly, 3- top wide-angle camera, 4- lidar, 5- front camera. DETAILED DESCRIPTION

[0021] The present invention is further described in detail below in conjunction with specific embodiments, which are intended to explain the present invention rather than to limit it.

[0022] like Figure 1 As shown, the present invention provides a vehicle bottom security inspection robot, including a carrier and a linear array camera assembly 1 installed on the carrier, a top ranging assembly 2, a top wide-angle camera 3, a laser radar 4, a front camera 5, a main control board and a data transmission module. The vehicle chassis security inspection robot integrates sensors such as linear array cameras, cameras, and laser radars. Through the laser radar and camera fusion algorithm, it can autonomously aim at the vehicle and enter the bottom of the vehicle to check whether there are any security inspection equipment illegally hidden or modified on the vehicle chassis. The carrier adopts a four-wheel differential model chassis, which can realize forward, backward, and differential rotation in situ at any angle.

[0023] The chassis image and video data are collected through sensors such as linear array cameras, cameras, and lidars. The large amount of collected data is compressed and encoded, and the encoded data is transmitted to the user-side app through the Wi-Fi data transmission module.

[0024] The vehicle driving system uses a motor driver or other power control equipment to convert the planned motion instructions into control signals such as the motor speed and steering to achieve the forward, backward, and turning movements of the under-vehicle security inspection robot. Encoders, gyroscopes and other sensors are used to obtain the position, speed and posture information of the under-vehicle security inspection robot in real time, and then compared with the planned path for closed-loop control and timely adjustment of movement to ensure the accuracy and stability of driving.

[0025] The user-side app decompresses, decodes, and analyzes the returned data to extract useful information.

[0026] In order to achieve more efficient vehicle chassis inspection, the present invention also provides a method for automatically aligning an underbody inspection robot with a vehicle, which can automatically identify and locate the vehicle to be inspected after startup, ensuring that the robot can accurately align with the chassis of the target vehicle, thereby performing an effective inspection.

[0027] The method for automatically aligning a vehicle with a vehicle bottom security inspection robot of the present invention comprises the following steps: Step 1: After the robot is started, it receives the real-time video stream from the front camera 5, performs image processing internally, runs the target detection algorithm on the obtained data, and identifies all the vehicles in the image. Each identified vehicle is marked with a rectangular box (x1, y1, x2, y2); Step 2: The video stream data of the front camera 5 is transmitted to the app via a wireless network; Step 3: After the user clicks on the vehicle to be tested on the screen in the app display interface, the click coordinate c1 will be sent to the robot. Based on the user's click coordinate c1, the robot program determines which rectangular box is the vehicle selected by the operator, and then calculates the lateral position range of the vehicle in the image, expressed as c2 = (xl, xr), which is the horizontal coordinates of the left and right edges of the vehicle on the image; Step 4: Convert the c2 value in the image coordinate system to the actual physical coordinate system of the laser radar 4. After completing the coordinate conversion, calculate the corresponding laser radar 4 scanning angle (a1, a2); Step 5: The laser radar 4 scans according to the starting angle a1 and the ending angle a2, searches for the left and right wheels of the vehicle, and obtains their coordinate positions c3 and c4; the position of the vehicle to be tested is inferred from the positions c3 and c4, and after obtaining the position of the vehicle to be tested, automatic alignment is performed to start the underbody scanning task.

[0028] The process of automatic alignment is: Step 1: When the robot receives the double-click command, it enters the automatic vehicle alignment mode; Step 2: The robot scans the front wheel track of the current vehicle through the laser radar 4, and selects the center axis of the two-wheel connection line to find the target point 0.5m away from the robot side; Step 3: Adjust the extension line of the robot's head to align with the target point and move to the target point; Step 4: Adjust the robot angle on the spot so that its body is completely aligned with the mid-perpendicular line of the target path.

[0029] The user side constructs the parking lot model through the app. The optional detection modes include: single car mode, longitudinal car mode, horizontal car mode, and multi-row custom mode.

[0030] The navigation method of the vehicle underbody inspection robot for autonomous batch inspection of parking lot vehicles of the present invention comprises the following contents: (1) Longitudinal train mode: like Figure 2 As shown, the operating app makes the underbody inspection robot pass through all the vehicles arranged in vertical columns in a straight line in sequence, and divides the image collected by the linear array camera component 1 according to the change in the distance between the robot and the bottom of the vehicle obtained by the robot top ranging component 2, and obtains the underbody pictures of each vehicle in sequence for grouping, storage and output.

[0031] like Figure 3 As shown, the specific method for batch inspection of vehicles in longitudinal train mode is: 1) The user inputs the total number of vehicles in the longitudinal train, and operates the app to move the undercarriage inspection robot to within a 45-degree angle range in the vertical direction of the front of the first vehicle in the longitudinal train. When the front camera 5 identifies the first vehicle to be inspected through the target recognition algorithm, the laser radar 4 is used to scan the front wheel track of the first vehicle, and the first target point is found 0.5m away from the center axis of the double-wheel connection line to the robot side; 2) The under-vehicle inspection robot reaches the first target point, so that the wheel tracks on both sides of the laser radar are the same. The front camera 5 recognizes the license plate information and the robot moves forward and passes under the first vehicle. When the distance measurement component 2 on the top of the vehicle detects that a vehicle has been inspected based on the distance change, the robot obtains the complete under-vehicle image captured by the linear array camera component 1 and issues a stop command to the robot to stop. 3) When the total number of vehicles in the longitudinal train is reduced by one, the laser radar 4 is aimed at the next vehicle again in a similar way to identify the license plate information. Regardless of whether the license plate is found or not, the underbody inspection robot scans the wheelbase in front through the laser radar 4 and calibrates the vehicle again. After the calibration is completed, the next vehicle is inspected until all vehicles are completely inspected and the underbody pictures of each vehicle are obtained.

[0032] (2) Horizontal car mode: like Figure 4 As shown, on the app side, the user independently builds a batch inspection model for vehicles parked in a row. The user operates the app to move the underbody inspection robot to within a 45-degree angle range in the vertical direction of the front of the first vehicle, and then starts a task of inspecting vehicles parked in a row. Each vehicle is inspected in turn until the underbody pictures of all vehicles are obtained.

[0033] like Figure 5 As shown, in this mode, users are required to input the width of the horizontal parking spaces and the information of whether the horizontal parking spaces are fully parked. On the app side, users can independently build a horizontal parking lot batch inspection model. Users first input the number of horizontal parking spaces and the width of the parking spaces, and the application side will display the corresponding horizontal parking lot model; then users need to manually mark the empty parking spaces, and then the application side will automatically generate the task model.

[0034] 1) The robot's front camera 5 identifies the vehicle to be detected through the target recognition algorithm; 2) The laser radar scans the front wheel track and selects the first target point 3m away from the center axis of the two-wheel connection to the robot side; 3) Autonomously navigate to the first target point, rotate the robot in place so that the wheelbases of both sides are the same as the laser radar, and then aim at the first vehicle and recognize the license plate; the robot moves forward and passes under the first vehicle. When the top distance measurement component determines that the robot has completely passed under the vehicle, the system obtains a complete picture of the vehicle bottom and then retreats back along the original route; 4) The robot returns to the position aimed at the first vehicle and determines whether all vehicle detections are completed. If completed, jump to step 6). If not completed, execute step 5); 5) Turn 90 degrees to the side of the next vehicle, and then drive forward a fixed distance D, where the fixed distance D is calculated by the formula Calculated, where W represents the width of the parking space, n represents the number of empty parking spaces. After reaching the target point, immediately rotate 90 degrees in the opposite direction and loop step 1) to step 4); 6) End the task and output and store all vehicle license plates and complete chassis image information.

[0035] (3) Multi-row custom mode: like Figure 6 As shown, the user can select the multi-row mode on the application side. When the multi-row mode is executed, select one of the longitudinal train mode or the transverse vehicle mode to execute the single-row mode. After executing the single-row, rotate 180 degrees and complete the scanning of the opposite row of vehicles in reverse, end the task, and output and store all vehicle license plates and complete chassis picture information.

Claims

1. A method for automatically aligning a vehicle with an under-vehicle security inspection robot, characterized in that: The following steps are involved: Step 1: After the robot starts, it receives the real-time video stream from the front camera and performs image processing internally. It runs the target detection algorithm on the obtained data to identify all the vehicles in the image. Each identified vehicle is marked with a rectangular box (x1, y1, x2, y2); Step 2: The video stream data of the front camera is transmitted to the app via a wireless network; Step 3: After the user clicks on the vehicle to be tested on the screen in the app display interface, the click coordinate c1 will be sent to the robot. Based on the user's click coordinate c1, the robot program determines which rectangular box is the vehicle selected by the operator, and then calculates the lateral position range of the vehicle in the image, expressed as c2 = (xl, xr), which is the horizontal coordinates of the left and right edges of the vehicle on the image; Step 4: Convert the c2 value in the image coordinate system to the actual physical coordinate system of the laser radar. After completing the coordinate conversion, calculate the corresponding angle interval (a1, a2) that the laser radar should scan. Step 5: The laser radar scans according to the starting angle a1 and the ending angle a2, searches for the left and right wheels of the vehicle, and obtains their coordinate positions c3 and c4; the position of the vehicle to be tested is inferred from the positions c3 and c4, and after obtaining the position of the vehicle to be tested, automatic alignment is performed to start the underbody scanning task.

2. The method for automatically aligning a vehicle according to claim 1, characterized in that: The process of automatic alignment described in step 5 is as follows: Step 1: When the robot receives the double-click command, it enters the automatic vehicle alignment mode; Step 2: The robot scans the front wheel track of the current vehicle through the laser radar, and selects the center axis of the two-wheel connection line to find the target point 0.5m away from the robot side; Step 3: Adjust the extension line of the robot's head to align with the target point and move to the target point; Step 4: Adjust the robot angle on the spot so that its body is completely aligned with the mid-perpendicular line of the target path.

3. An autonomous navigation batch inspection vehicle method based on the automatic vehicle alignment method of claim 2, characterized in that: The parking lot model and detection model are built through the user-side app. The detection modes include longitudinal train mode, transverse car mode and multi-row custom mode. Train mode: S1.

1. The user inputs the total number of vehicles in the longitudinal train, and operates the app to move the undercarriage inspection robot to the front position of the first vehicle in the longitudinal train. When the front camera identifies the first vehicle to be inspected through the target recognition algorithm, the laser radar is used to scan the front wheel track of the first vehicle, and the center axis of the double-wheel connection line is selected to find the first target point 0.5m away from the robot side; S1.2, the under-vehicle security inspection robot reaches the first target point, so that the wheel tracks on both sides of the laser radar are the same, and the license plate information is recognized by the front camera. The robot moves forward and passes under the first vehicle. When the distance measurement component on the top of the vehicle detects that a vehicle has been inspected according to the distance change, the complete under-vehicle image captured by the linear array camera component is obtained, and a stop command is issued to the robot to stop; S1.3, the total number of vehicles in the longitudinal train is reduced by one, and the laser radar is again aimed at the next vehicle in a similar way to identify the license plate information. Regardless of whether the license plate is found, the undercarriage inspection robot scans the wheelbase in front through the laser radar and calibrates the vehicle again. After the calibration is completed, the next vehicle is inspected until all vehicles are completely inspected and the undercarriage images of each vehicle are obtained; Horizontal car mode: S2.

1. Use the app to move the undercarriage inspection robot to the front of the first car in the longitudinal train. The robot's front camera (5) recognizes the vehicle to be inspected through the target recognition algorithm. S2.

2. Scan the front wheel track with the laser radar and find the first target point 3m away from the center axis of the two-wheel connection line to the robot side. S2.

3. Autonomously navigate to the first target point, rotate the robot in place so that the wheelbases of both sides of the laser radar are the same. At this time, the robot aims at the first vehicle and recognizes the license plate. The robot moves forward and passes under the first vehicle. When the top distance measurement component determines that the robot has completely passed under the vehicle, the system obtains a complete picture of the vehicle bottom and then retreats back along the original route. S2.4, the robot returns to the posture aimed at the first vehicle and determines whether all vehicle inspections have been completed. If completed, it jumps to step S2.6; if not, it executes step S2.5; S2.5, turn 90 degrees to the side of the next vehicle, and then drive forward a fixed distance D, where the fixed distance D is calculated by the formula Calculated, where W represents the width of the parking space, n represents the number of empty parking spaces. After reaching the target point, immediately rotate 90 degrees in the opposite direction and loop through steps S2.1 to S2.4; S2.6, end the task, output and store all vehicle license plates and complete chassis image information; Multi-row custom mode: When executing the multi-row mode, select one of the longitudinal train mode or the transverse vehicle mode to execute the single-row mode. After executing the single-row, rotate 180 degrees and complete the scanning of the opposite row of vehicles in reverse, end the task, and output and store all vehicle license plates and complete chassis picture information.

4. The autonomous navigation batch inspection vehicle method according to claim 3 is characterized in that: In the horizontal parking mode, the user first enters the number of horizontal parking spaces, parking space width information, marks empty parking spaces, and then generates a horizontal parking batch inspection model.

Images