A system and method for automatically collecting and detecting collision conditions of shared cars

Abstract

The present invention discloses a system and method for automatically collecting and detecting collision conditions of shared cars. The system collects full-car pictures before and after the use of the vehicle through two retractable cameras installed at the front and rear license plate positions and two 360-degree panoramic cameras at the left and right rearview mirrors of the car body, and collects collision sounds through directional sound sensors around the chassis and body. The two pieces of information are combined to determine whether the car body has any collisions or scratches during the use of the shared car, and the information is uploaded to the shared car operation platform in real time. The system is preferentially applied to vehicles that can be shared. When a user rents a shared car, the vehicle controller must first perform a vehicle condition detection on the vehicle, store the vehicle condition detection results, and send them to the user client to confirm that the vehicle can be used. When the user finishes using the vehicle, the vehicle controller performs a vehicle condition detection on the vehicle again, and compares the vehicle condition detection results with the detection results before the user uses the vehicle, so as to ensure the user's safe use of the vehicle and reduce economic disputes.

Description

Technical Field

[0001] The present invention relates to the technical field of shared vehicles. More specifically, the present invention relates to a system and method for automatically collecting and detecting collision conditions of shared vehicles. Background Art

[0002] With the increasing number of cars in my country and the cost of using them, more and more people are choosing shared cars as a means of transportation. With the continuous advancement of electrification, connectivity, intelligence, and sharing, car sharing has greatly facilitated people's lives. However, due to the public nature of shared cars, their use and management also present numerous challenges. For users, to avoid erroneous attribution of responsibility, drivers generally err on the side of caution and record the condition of their shared cars before and after each use for future reference. This significantly reduces the convenience of shared car use. For shared car operators, this also requires frequent visits by dedicated personnel to inspect and verify any damage to shared cars during use or while they are parked, to determine whether the damage was caused by the user. This significantly increases labor costs for operators due to the extensive manpower input, as well as the difficulties in determining liability and subsequent financial losses caused by non-first-hand damage assessment methods. Therefore, it is worthwhile to explore how to automatically collect and detect damage to cars during use and while they are parked, significantly reducing the management costs and financial disputes associated with existing shared car operations.

[0003] In order to overcome the problems of inaccurate liability determination and high costs caused by the need to manually record the collision conditions of their cars during user use, shared car operators have now developed some automatic vehicle condition monitoring devices.

[0004] For example, the utility model patent "Shared Car Condition Automatic Monitoring and Uploading System" (authorization publication number: CN209000079U) discloses a shared car condition automatic monitoring and uploading system, which uses magnetostrictive displacement sensors and cameras to collect body collision signals and vehicle conditions. This automatic vehicle condition monitoring system, on the one hand, is unable to fully collect the entire vehicle condition and identify vehicle scratches due to the low detection accuracy of the sensors and the blind spots in the camera layout. On the other hand, because magnetostrictive displacement sensors are required to be applied to the entire vehicle body to detect surface damage, there are problems such as an excessive number of sensors, high costs, and inconvenience in vehicle maintenance, making it of no practical application significance.

[0005] In addition, the invention patent "A method for verifying the condition of a shared driverless car when renting it" (authorization publication number: CN1707813775U) discloses a shared driverless car inspection platform that uses multiple cameras to collect information about the inside and outside conditions of the car. The unmanned shared car is then dispatched to the user's location, and the user takes a photo to confirm the condition of the car again. The solution proposed in this patent is a method for inspecting the condition of shared cars through an inspection platform. Since the shared car must arrive at a designated location for inspection, the real-time and convenient vehicle condition inspection requirements cannot be achieved, and the problem of accurately allocating liability for damage to the user's car cannot be solved.

[0006] On the one hand, the above-mentioned vehicle inspection platforms cannot perform real-time vehicle condition detection, which requires users to intervene too much in the inspection process; on the other hand, these methods only involve vehicle condition detection before users rent a vehicle, and cannot cover the responsibility judgment of the entire vehicle rental process, and therefore do not have high application value. Summary of the Invention

[0007] The purpose of the present invention is to provide a system and method for automatically collecting and detecting the collision conditions of shared cars. The system uses a camera and a directional sound sensor installed on the car body to automatically collect and judge the shared car condition information, and automatically detect and identify various collisions on the car body, including scratches, collisions and damages, etc., to provide convenience for drivers during use and reduce unnecessary economic disputes between operators and users due to accident damage assessment and responsibility division, while improving the efficiency of shared car company operations and management and reducing costs.

[0008] Beneficial effects of the present invention:

[0009] 1. The present invention can minimize the asset losses of shared car operating companies and insurance companies, and reduce economic disputes between operating companies and users caused by damage to shared cars. The technical solution adopted by the present invention adopts an on-board all-round self-inspection system for vehicle body damage, which can accurately detect vehicle damage during the user's driving and parking period. It can not only eliminate the cumbersome experience of users taking their own photos for evidence, but also combine the on-board automatic detection of shared cars with uploading to the cloud platform to compare the damage situation, and more accurately divide the responsibility for vehicle damage, avoiding the problem of users' underreporting, concealing the damaged vehicle condition, and misidentifying the responsibility.

[0010] 2. The present invention automatically collects vehicle condition information without blind spots. The system and method for automatically collecting and detecting collisions in shared vehicles, described in this invention, recognizes user rental / return signals and triggers the operation of cameras with automatically retractable rods mounted between the front and rear bumpers and cameras below the left and right rearview mirrors. This automatically collects information about vehicle damage during the handover process.

[0011] 3. The present invention employs diverse technical means for detecting vehicle conditions. The method for automatically collecting and detecting collisions in shared vehicles, described in the present invention, combines image comparison with directional sound sensor signals to effectively identify various collisions, including scratches, collisions, and damage, in specific locations on the vehicle body while it is in motion. It also introduces similarity method parameters to quantify the degree of similarity between vehicle condition images before and after a vehicle is handed over.

[0012] 4. The detection device of the present invention has low installation costs and high reliability. The detection equipment used in the automatic collection and detection system for shared car collision conditions described in the present invention is a camera and directional sound sensor commonly used on existing vehicles. This has low manufacturing costs and requires minimal modification to the original vehicle. Furthermore, its telescopic structure allows it to be stowed into the vehicle while driving, without changing the vehicle's appearance, improving installation reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 This is a block diagram of the system for automatically collecting and detecting collision conditions of shared cars according to the present invention;

[0014] Figure 2 This is a diagram showing the layout of the components of the automatic collection and detection system for collision conditions of shared vehicles described in the present invention;

[0015] Figure 3 This is a general control flow chart of a method for automatically collecting and detecting collision conditions of shared vehicles according to the present invention;

[0016] Figure 4 This is a flowchart of a rental car inspection procedure for a method for automatically collecting and detecting collision conditions of shared cars according to the present invention;

[0017] Figure 5 This is a flowchart of a real-time monitoring program of a directional sound sensor for a method for automatically collecting and detecting collision conditions of shared vehicles according to the present invention;

[0018] Figure 6 This is a flow chart of the vehicle delivery and re-inspection procedure for the method for automatically collecting and detecting collision conditions of shared vehicles according to the present invention;

[0019] Figure 7 This is a flowchart of a vehicle condition photo similarity comparison program for a method for automatically collecting and detecting collision conditions of shared vehicles described in the present invention;

[0020] Figure 8 This is a flowchart of the vehicle body directional sound similarity comparison program for the method for automatically collecting and detecting collision conditions of shared vehicles described in the present invention;

[0021] Figure 9This is a flowchart of the chassis directional sound similarity comparison program for the method for automatically collecting and detecting collision conditions of shared cars described in the present invention;

[0022] Figure 10 This is a flowchart of the startup procedure of the automatic retractable device for the automatic collection and detection system of collision conditions of shared cars according to the present invention;

[0023] Figure 11 This is a simplified working diagram of the automatic telescopic device for the automatic collection and detection system of collision conditions of shared cars described in the present invention. DETAILED DESCRIPTION

[0024] The present invention will be described in further detail below in conjunction with the accompanying drawings so that those skilled in the art can implement the invention with reference to the description.

[0025] It should be understood that terms such as “having,” “including,” and “comprising” used herein do not prescribe the existence or addition of one or more other elements or combinations thereof.

[0026] like Figure 1As shown, the present invention shows a system for automatically collecting and detecting collision conditions of shared cars, which includes six directional sound sensors, four on-board cameras, a vehicle condition information recognition controller, an automatic telescopic device, an ultrasonic radar, and the input and receiving ends of the 5G operation background. The six directional sound sensors include a front directional sound sensor, a rear directional sound sensor, a left directional sound sensor, a right directional sound sensor, a chassis front directional sound sensor, and a chassis rear directional sound sensor. The four on-board cameras include a front camera, a rear camera, a left on-board camera, and a right on-board camera. The ultrasonic radar, directional sound sensor, and on-board camera are respectively connected to the input ends of the A / D converter on the vehicle condition information recognition controller. The collected front and rear distance information, sound sensor signals, and car photo information are transmitted to the A / D converter on the vehicle condition information recognition controller in the form of analog signals. The A / D converter converts the analog signals into digital signals so that the controller main chip can compare the vehicle condition information. The vehicle condition information recognition controller's main chip is connected to the input of its built-in D / A converter. Before the vehicle condition information comparison program runs, the controller sends a digital control signal, which is converted into an analog control signal via the D / A converter. This signal then controls the automatic retractable mechanism to execute the corresponding movement and activate the camera. Furthermore, the vehicle condition information recognition controller communicates with the operations backend via the 5G network, retrieving initial sensor signals and vehicle photo information from the 5G operations backend's transmitter. After the comparison program is complete, the updated sensor signals, vehicle photo information, and comparison results are uploaded to the 5G operations backend's receiver. The vehicle condition information recognition controller can also retrieve vehicle status information reflecting the current vehicle usage, such as vehicle speed, engine speed, and brake pedal opening, from the body controller via the CAN bus to determine whether the vehicle's usage status is consistent with the current program execution stage.

[0027] like Figure 2 As shown in the layout diagram, the shared car is mainly equipped with an automatic telescopic device and a front camera 1, a vehicle condition information recognition controller 2, a body controller 3, a left-side vehicle-mounted camera 4, a right-side vehicle-mounted camera 5, an automatic telescopic device and a rear camera 6, a front-end directional sound sensor 7, a rear-end directional sound sensor 8, a left-side directional sound sensor 9, a right-side directional sound sensor 10, a chassis front-end directional sound sensor 11, and a chassis rear-end directional sound sensor 12.

[0028] Among them: the automatic telescopic device is designed to be set up in the front and rear of the car, and is fixedly connected to the front camera and the rear camera respectively, wherein the automatic telescopic device and the front camera 1 are arranged in the middle position of the bottom end of the front bumper of the car, and the automatic telescopic device and the rear camera 6 are arranged in the middle position of the bottom end of the rear bumper of the car; the automatic telescopic device specifically includes: a telescopic rod, a telescopic control motor, and a rotation control motor; the end turntable of the telescopic rod is fixed with the front camera or the rear camera, and the telescopic control motor can control the telescopic rod to change the distance between the end turntable and the car body, thereby ensuring the distance between the front camera or the rear camera and the car body, so as to ensure that it does not collide with surrounding obstacles when extended while effectively photographing the damage to the front and rear car body surfaces; the rotation control motor can control the rotation of the end turntable to ensure that the front camera or the rear camera is facing the car body, so as to facilitate complete and clear photography of the damage to the front and rear car body surfaces. The vehicle condition information recognition controller 2 is located in the instrument panel; the body controller 3 is arranged inside the vehicle control panel; the left vehicle-mounted camera 4 is arranged at the lower end of the left rearview mirror of the vehicle; the right vehicle-mounted camera 5 is arranged at the lower end of the right rearview mirror of the vehicle; the front end directional sound sensor 7 is arranged on the inside of the front bumper; the rear end directional sound sensor 8 is arranged on the inside of the rear bumper; the left side directional sound sensor 9 is arranged on the lower edge of the inside of the left door panel of the rear seat of the vehicle; the right side directional sound sensor 10 is arranged on the lower edge of the inside of the right door panel of the rear seat of the vehicle; the chassis front end directional sound sensor 11 is located near the front suspension subframe of the chassis; the chassis rear end directional sound sensor 12 is located near the rear suspension subframe of the chassis.

[0029] The vehicle condition information recognition controller of the present invention internally stores a method for automatically collecting and detecting collision conditions of shared vehicles, such as Figure 3 As shown, the overall control process of the method for automatically collecting and detecting collision conditions of shared cars according to the present invention specifically executes the following steps:

[0030] Step 1: The vehicle condition information recognition controller continuously communicates with the cloud to detect whether a cloud signal is transmitted;

[0031] Step 2: If the signal transmitted from the cloud is a stage signal (stage signals include parking and engine shutdown stage signals, driving monitoring stage signals, rental vehicle inspection stage signals, and delivery and re-inspection stage signals), the stage signal is temporarily stored in the vehicle condition information recognition controller;

[0032] Step 3: If the stage signal is a parking and engine-off stage signal or a driving monitoring stage signal, the vehicle condition information recognition controller will call the body and chassis directional sound similarity comparison program.

[0033] Step 4: If the signal at this stage is the rental vehicle inspection stage signal, the vehicle condition information recognition controller will call the rental vehicle inspection program and feed the results back to the cloud;

[0034] Step 5: Determine whether the current vehicle condition is a high-discrepancy condition based on the return value of the rental vehicle inspection program. If the current vehicle condition is a high-discrepancy condition, feedback is sent to the cloud, which then sends the vehicle damage information to the user's mobile app and asks the user whether to continue renting the vehicle. If it is not a high-discrepancy condition, proceed to the next step;

[0035] Step 6: Determine whether the current vehicle condition is medium or poor based on the program return value. If so, proceed to the next step. If not, determine that the vehicle is in low or poor condition, which is good and meets the rental conditions, and send a rental permission signal to the cloud.

[0036] Step 7: Determine whether the directional sound sensor's real-time monitoring program results are highly similar. If so, the vehicle is determined to have been bumped or scratched, and the report is fed back to the cloud, which then asks the user whether to continue renting the vehicle. If not, proceed to the next step.

[0037] Step 8: Determine whether the directional sound sensor's real-time monitoring program results show a medium degree of similarity. If so, upload the images taken by the camera during the previous vehicle delivery re-inspection phase to the cloud for manual review. If not, the vehicle is determined to be in good condition and meets the rental conditions, and a rental permission signal is sent to the cloud.

[0038] Step 9: If the signal at this stage is the vehicle delivery re-inspection stage signal, the vehicle condition information recognition controller will call the vehicle delivery re-inspection program and feedback the results to the cloud;

[0039] Step 10: Determine whether the current vehicle condition is a high-discrepancy condition based on the program's return value. If the current vehicle condition is a high-discrepancy condition, feedback the damaged vehicle condition information to the cloud, which then sends the vehicle damage information to the user's mobile phone APP and conducts corresponding accountability. If it is not a high-discrepancy condition, proceed to the next step;

[0040] Step 11: Determine whether the current vehicle condition is medium or poor based on the program return value. If so, proceed to the next step. If not, determine that the vehicle is in low or poor condition, and is in good condition, meeting the return conditions. A return permission signal is then sent to the cloud.

[0041] Step 12: Determine whether the result returned by the directional sound sensor real-time monitoring program is highly similar. If it is highly similar, determine that the vehicle has been bumped or scratched, and report it to the cloud for accountability. If it is not highly similar, proceed to the next step;

[0042] Step 13: Determine whether the result returned by the directional sound sensor real-time monitoring program is of medium similarity. If it is, upload the image taken by the camera during the previous vehicle delivery re-inspection phase to the cloud for manual judgment. If it is not of medium similarity, the vehicle is determined to be in good condition and meets the rental conditions, and a signal is sent to the cloud to allow the vehicle to be returned;

[0043] Step 14: The program returns to step 1.

[0044] like Figure 4 As shown, the specific execution steps of the car rental inspection program for the automatic collection and detection method of shared car collision conditions described in the present invention are as follows:

[0045] Step 1: To start, the user unlocks the rental car by scanning the QR code on the mobile client, and the controller recognizes the rental car signal;

[0046] Step 2: Determine whether the vehicle status stored in the body controller is consistent with the phase signal status fed back by the cloud;

[0047] Step 3: If the two are consistent, the cloud platform sends the vehicle status information uploaded to the cloud at the end of the previous rental via 5G network signals, that is, the car image captured by the camera;

[0048] Step 4: The vehicle status information recognition controller reads the vehicle status information sent by the cloud platform and temporarily stores it inside the controller. At the same time, the controller calls the automatic retractable device startup program;

[0049] Step 5: After the automatic retractable device startup program is completed, the vehicle condition photo similarity comparison program is called to perform a similarity comparison on the set of images to determine whether the user's driving process has caused damage to the vehicle;

[0050] Step 6: Return the result to the main program.

[0051] like Figure 5 As shown, the specific execution steps of the directional sound sensor real-time monitoring program for the automatic collection and detection method of shared car collision conditions described in the present invention are as follows:

[0052] Step 1: Read the vehicle status information stored in the body controller;

[0053] Step 2: Determine whether the vehicle status information is consistent with the program stage information. If the judgment result is no, then end the program and return the corresponding result. If the judgment result is yes, then proceed to the next step;

[0054] Step 3: Receive signals from six directional sound sensors and call the vehicle body directional sound similarity comparison program and the chassis directional sound sensor comparison program respectively for real-time judgment;

[0055] Step 4: Determine whether the comparison result is abnormal. If yes, store the position of the sensor corresponding to the abnormal sound signal and the similarity calculated by the comparison program in the vehicle condition information recognition controller and proceed to the next step. If no, proceed directly to the next step;

[0056] Step 5: Check whether the program phase information stored in the vehicle condition information recognition controller has changed. If not, continue to step 3. If yes, end the program.

[0057] like Figure 6 As shown, the specific execution steps of the vehicle delivery re-inspection procedure for the method for automatically collecting and detecting collision conditions of shared vehicles described in the present invention are as follows:

[0058] Step 1: The user sends a vehicle return command to the cloud through the mobile client, and the vehicle condition information recognition controller recognizes the vehicle return signal;

[0059] Step 2: Determine whether the vehicle status stored in the body controller is consistent with the phase signal status fed back by the cloud;

[0060] Step 3: If the two are consistent, the cloud platform sends the vehicle status information uploaded to the cloud at the beginning of the rental via 5G network signals, that is, the car image captured by the camera;

[0061] Step 4: The vehicle status information recognition controller reads the vehicle status information sent by the cloud platform and temporarily stores it internally. At the same time, the controller calls the automatic retractable device startup program;

[0062] Step 5: After the automatic detection process is completed, the vehicle condition photo similarity comparison program is called to perform a similarity comparison on the set of images to determine whether the user's driving process has caused damage to the vehicle condition;

[0063] Step 6: Return the result to the main program.

[0064] like Figure 7 As shown, the specific execution steps of the vehicle condition photo similarity comparison program for the method for automatically collecting and detecting collision conditions of shared vehicles described in the present invention are as follows:

[0065] Step 1: The vehicle condition information recognition controller calls out the two most recently updated vehicle condition photos stored in the controller to form a control group for subsequent image comparison;

[0066] Step 2: Determine the similarity parameters of the group of photos using a comparison program and record them in the controller;

[0067] Step 3: The comparison program sets a similarity upper limit S 上 and the similarity lower limit S 下 Two numerical values;

[0068] 1. When the similarity is between 0 and S 下 When the vehicle is in the state of collision or scratch in the previous stage, the vehicle condition at this time is defined as a high difference vehicle condition.

[0069] 2. When the similarity is between S 下 and S 上 When the vehicle condition is between 0 and 1, the vehicle condition at this time is defined as a medium difference vehicle condition, and further comprehensive analysis of the vehicle condition at this time is required based on the results obtained by the directional sound sensor;

[0070] 3. When the similarity is greater than S 上 When , the car is defined as undamaged;

[0071] Step 4: If medium or high difference in vehicle conditions occurs, the image information will be fed back to the cloud via the 5G network to inform the platform personnel of the damage to the vehicle. The cloud platform will then inform the user through the user's mobile client APP or negotiate with the user on compensation matters.

[0072] like Figure 8 As shown, the specific execution steps of the vehicle body directional sound similarity comparison program for the method for automatically collecting and detecting collision conditions of shared vehicles described in the present invention are as follows:

[0073] Step 1: Receive signals from four vehicle body directional sound sensors;

[0074] Step 2: The sound intensity S of the directional sound sensor signals received in real time at the front, rear, left and right sides of the car n , n = 1, 2, 3, 4 are respectively compared with the sound intensity S0 of the directional sound sensor signal when the vehicle body is bumped in the pre-stored state to obtain the corresponding similarity M n , n=1,2,3,4, that is

[0075] Step 3: Compare and get the corresponding similarity M n ,n=1,2,3,4 respectively with M 下 Compare and judge M n >M 下 Is it true? 下 is the similarity threshold for no scratches or bumps, and its specific value can be determined experimentally. If the judgment result of the similarity corresponding to a certain sensor is negative, the similarity is judged to be low, and the car in the direction corresponding to the sensor has not experienced any scratches or bumps, and the process continues to step 6. If the judgment result is positive, the similarity continues to participate in the next step of the process;

[0076] Step 4: Match the judgment formula M in step 3 n >M 下 The similarity M n, respectively with M 上 Compare and judge M n >M 上 Is it true? 上 The threshold for similarity that indicates only scratches occurred. The specific value can be determined experimentally. If the judgment result of a sensor's corresponding similarity is negative, the similarity is considered medium, and the car in the direction corresponding to the sensor only experienced scratches. If the judgment result is positive, the similarity is considered high, and the car in the direction corresponding to the sensor experienced both scratches and collisions.

[0077] Step 5: Return to the position corresponding to the directional sound sensor in step 4;

[0078] Step 6: Return the corresponding result and the program ends.

[0079] like Figure 9 As shown, the specific execution steps of the chassis directional sound similarity comparison program for the automatic collection and detection method of shared car collision conditions described in the present invention are as follows:

[0080] Step 1: Receive signals from two chassis directional sound sensors;

[0081] Step 2: The sound intensity S of the directional sound sensor signal received in real time at the front and rear ends of the chassis n , n = 5, 6 are compared with the sound intensity S′0 of the directional sound sensor signal when the chassis is bumped in advance to obtain the corresponding similarity M n , n=5,6, that is

[0082] Step 3: Compare and get the corresponding similarity M n , n=5,6 respectively with M 安全 Compare and judge M n >M 安全 Is it true? 安全 The threshold value for similarity that the chassis has not suffered serious scratches or bumps is the similarity threshold, and its specific value can be determined by experiment. If the judgment result of the corresponding similarity of a certain sensor is no, then this similarity is judged to be safe similarity, and the chassis of the car in the direction corresponding to this sensor has not suffered serious scratches or bumps. If the judgment result of the corresponding similarity of a certain directional sound sensor is yes, then this similarity is judged to be dangerous similarity, and the chassis of the car in the direction corresponding to this directional sound sensor has suffered serious scratches or bumps, and the corresponding position of the directional sound sensor is returned to the cloud;

[0083] Step 4: Return the corresponding result of step 3 to the cloud. The cloud records the dangerous similarity and the rental user information for subsequent maintenance and accountability. This process ends.

[0084] like Figure 10As shown, the vehicle condition information recognition controller of the present invention also stores a program for starting the automatic telescopic device when renting and delivering the vehicle. The specific steps of the program are as follows:

[0085] Step 1: The ultrasonic radar equipped in the housing of the device measures the straight-line distance between the device and the obstacle in front, which is recorded as L1, and the straight-line distance between the device and the obstacle behind, which is recorded as L2, and the data is stored in the vehicle condition information recognition controller;

[0086] Step 2: Based on the straight-line distances L1 and L2, the controller can calculate the telescopic distances D1 and D2 of the telescopic rods in the front and rear automatic telescopic devices respectively, and determine whether D1 and D2 have reached the minimum distance. If the minimum distance is not reached, the controller will feedback to the user to adjust the vehicle position. After the user adjusts, the program will be triggered again. If the user cannot find a suitable position, they can choose to manually take a photo and upload it. If the minimum distance is exceeded, the corresponding control signals based on D1 and D2 will be transmitted to the motors T1 and T2 respectively;

[0087] Step 3: After receiving the signal, motors T1 and T2 drive their respective telescopic rods to reach telescopic distances D1 and D2 respectively;

[0088] Step 4: Based on the straight-line distances L1 and L2 and other known quantities, the controller can calculate the first rotation angles γ1 and γ2 of the front and rear telescopic cameras, and transmit the corresponding control signals to motors T′1 and T′2. Motors T′1 and T′2 rotate the corresponding angles to drive their respective cameras to the appropriate position. At the same time, the front and rear telescopic cameras, left and right cameras are activated to take the first photo, and the collected vehicle condition photos are stored in the controller;

[0089] Step 5: Based on the straight-line distances L1 and L2 and other known quantities, the controller can calculate the second rotation angles γ′1 and γ′2 of the front camera and the rear telescopic camera, and transmit the corresponding control signals to the motors T′1 and T′2. The motors T′1 and T′2 rotate the corresponding angles to drive their respective cameras to the appropriate position. At the same time, the front telescopic camera, the rear telescopic camera, the left camera, and the right camera are activated to take a second photo, and the collected vehicle condition photos are stored in the controller;

[0090] Step 6: After successfully completing the above steps, the front telescopic camera and the rear telescopic camera are rotated in the opposite direction by angles γ″1 and γ″2 respectively, returning to their respective initial positions;

[0091] Step 7: Motors T1 and T2 drive the telescopic rod to retract to the initial position, and the program ends.

[0092] The motion control method of the automatic telescopic device is as follows: the automatic telescopic device will trigger the automatic telescopic device startup program when renting and delivering the vehicle. The vehicle condition information recognition controller stores the telescopic movement distance D of the front and rear automatic telescopic devices. n The first upward rotation angle of the camera fixed at the end is γ n and the subsequent second downward rotation angle γ′ n Here, we use the rear-mounted automatic telescopic device as an example to explain how to calculate the telescopic distance and the camera's two rotation angles.

[0093] like Figure 11 As shown, point A and point F are the lowest points of the front and rear bumpers respectively, point B and point E are the starting points of the telescopic rods of the front and rear automatic telescopic devices, point C and point D are the vertices of the front and rear windows respectively, and point G is the spatial position point where the camera finally stops moving to shoot the car body.

[0094] The calculation formula for the telescopic movement distance D of the automatic telescopic device is as follows:

[0095]

[0096] In the above formula, L is the straight-line distance between the end of the telescopic rod of the rear automatic telescopic device and the rear obstacle detected by the ultrasonic radar distance sensor; α is the fixed angle formed by the telescopic rod and the horizontal plane, which is 45° in this embodiment; S is the spacing margin reserved for safety, which is generally 0.1 to 0.2 m.

[0097] H=(LS)×tanα

[0098] Where H is the vertical distance between point G and point B.

[0099] The calculation formula for the first upward rotation angle γ2 of the rear camera at the end of the automatic retractable device after it is extended is as follows:

[0100]

[0101] After the first rotation to take the photo, the second rotation is performed downwards. The calculation formula for the rotation angle γ′2 is as follows:

[0102]

[0103] Where, β1 is the angle between the line connecting points B and G and the vertical direction; β2 is the angle between the line connecting points A and G and the vertical direction; β3 is the angle between the line connecting points C and G and the vertical direction. is the effective shooting range angle of the camera in the vertical direction; M is the horizontal distance between points C and G; K is the vertical distance between points C and A; N is the vertical distance between points A and B.

Claims

1. A system for automatically collecting and detecting collision conditions of shared vehicles, characterized in that: include: Directional sound sensors, which are used to collect sound signals at corresponding positions of the car, include: front directional sound sensor, rear directional sound sensor, left directional sound sensor, right directional sound sensor, chassis front directional sound sensor, and chassis rear directional sound sensor; Front and rear telescopic cameras, including: front camera and rear camera, used to collect photos of the front and rear of the car respectively; The on-board cameras on both sides include: a left-side on-board camera and a right-side on-board camera, which are used to collect photo information of the left and right sides of the car respectively; Automatic telescopic devices, with two sets located on the front and rear bumpers of the vehicle, respectively, include: a telescopic rod, a telescopic control motor, and a rotation control motor; the ends of the telescopic rods are fixed to the front and rear telescopic cameras; the telescopic control motor controls the telescopic rods to extend and retract to change the distance between the front and rear telescopic cameras and the vehicle body; and the rotation control motor controls the front and rear telescopic cameras to rotate up and down to ensure that the viewing angle covers the front and rear ends of the vehicle body; Ultrasonic radar, used to measure the straight-line distance between the end of the telescopic rod and obstacles in front of and behind the car; The body controller is used to collect status information such as vehicle speed, engine speed, and brake pedal opening to reflect the vehicle's usage status; A vehicle condition information recognition controller internally stores a method for automatically collecting and detecting collision conditions of a shared vehicle. The controller determines whether the vehicle's usage status is consistent with the current program operation stage based on the sound, image, and distance signals detected by the directional sound sensor, the front and rear telescopic cameras, the two side vehicle-mounted cameras, and the ultrasonic radar, as well as vehicle status information read from the body controller via the bus. The controller also interacts with the cloud platform via the 5G network to achieve image collection, comparison, determination, and data communication of the vehicle's collision status, as well as motion control of the automatic telescopic device. The method for automatically collecting and detecting collision conditions of shared vehicles is characterized by comprising: An automated acquisition and detection program uses image comparison combined with directional sound sensor signals to effectively identify various collisions, including scratches, collisions, and damage, in specific locations on the vehicle while it is in motion. It also introduces similarity method parameters to quantify the degree of similarity between images of the vehicle before and after handover. A car rental inspection program, which automatically detects and identifies any scratches, collisions, and damage to the vehicle before the user rents it, and reports the results to the cloud for reference by the user and backend management staff. The vehicle delivery re-inspection program is used to automatically detect and identify various collisions, including scratches, collisions, and damage, on the vehicle body after the user takes possession of the vehicle and feeds the results back to the cloud for reference by the user and back-end management personnel; A vehicle condition photo similarity comparison program is used to compare the two most recently updated vehicle condition photos of a car, using similarity parameters to quantify the degree of difference in vehicle condition and thus determine the vehicle's collision condition; a directional sound sensor real-time monitoring program for reading directional sound sensor signals distributed on the vehicle body and chassis based on current vehicle status information, calling the vehicle body directional sound similarity comparison program and the chassis directional sound similarity comparison program, and determining whether an abnormal sound signal occurs and its direction; A vehicle body directional sound similarity comparison program is used to identify abnormal sound signals at corresponding locations on the vehicle body, thereby determining whether the vehicle body has been scratched, collided, or damaged; Chassis directional sound similarity comparison program, which is used to identify abnormal sound signals at corresponding locations on the chassis, and then determine whether the chassis is scratched, collided or damaged; An automatic telescopic device activation program is used to control the rotation of two telescopic drive motors of the front and rear automatic telescopic devices disposed on the front and rear bumpers of the vehicle to drive their telescopic rods to extend and retract, and to control the front and rear cameras fixed to the ends of the automatic telescopic devices to rotate up and down and start shooting under the rotation of the two rotation drive motors; The automatic collection and detection program includes: Step 1: The vehicle condition information recognition controller continuously communicates with the cloud to detect whether there is a signal transmitted from the cloud; Step 2: If the signal transmitted from the cloud is a phase signal, which includes a parking and engine shutdown phase signal, a driving monitoring phase signal, a rental vehicle inspection phase signal, and a delivery and re-inspection phase signal, the phase signal is temporarily stored in the vehicle condition information recognition controller; Step 3: If the stage signal is a parking and engine-off stage signal or a driving monitoring stage signal, the vehicle condition information recognition controller will call the directional sound sensor real-time monitoring program; Step 4: If the signal is a rental vehicle inspection signal, the vehicle condition information recognition controller will call the rental vehicle inspection program and feed the result back to the cloud; Step 5: Determine whether the current vehicle condition is a high-discrepancy condition based on the return value of the rental vehicle inspection program; if the current vehicle condition is a high-discrepancy condition, feedback the damaged vehicle condition information to the cloud, which then sends the vehicle damage information to the user's mobile phone app and asks the user whether to continue renting the vehicle; if the current vehicle condition is not a high-discrepancy condition, proceed to the next step; Step 6: Determine whether the current vehicle condition is medium or poor based on the program return value; if so, proceed to the next step; if not, determine that the vehicle is in low or poor condition, is in good condition, and meets the rental conditions, and send a rental permission signal to the cloud; Step 7: Determine whether the result returned by the directional sound sensor real-time monitoring program is highly similar; if so, determine that the vehicle has been bumped or scratched, and report this to the cloud, which then asks the user whether to continue renting the vehicle; if not, proceed to the next step; Step 8: Determine whether the result returned by the directional sound sensor real-time monitoring program is of medium similarity; if so, upload the image captured by the camera during the previous vehicle delivery re-inspection phase to the cloud for manual review; if not, determine that the vehicle is in good condition and meets the rental conditions, and send a rental permission signal to the cloud; Step 9: If the stage signal is a vehicle delivery re-inspection stage signal, the vehicle condition information recognition controller will call the vehicle delivery re-inspection program and feed the result back to the cloud; Step 10: Determine whether the current vehicle condition is a high-discrepancy condition based on the program's return value. If so, the damaged vehicle condition information is fed back to the cloud, which then sends the vehicle damage information to the user's mobile app and conducts corresponding accountability. If not, proceed to the next step. Step 11: Determine whether the current vehicle condition is a medium-differential condition based on the program return value; if the current vehicle condition is a medium-differential condition, proceed to the next step; if the current vehicle condition is not a medium-differential condition, determine that the current vehicle condition is a low-differential condition, the vehicle condition is good, and meets the return conditions, and send a return permission signal to the cloud; Step 12: Determine whether the result returned by the directional sound sensor real-time monitoring program is highly similar; if so, determine that the vehicle has been bumped or scratched, and report this to the cloud for accountability; if not, proceed to the next step; Step 13: Determine whether the result returned by the directional sound sensor real-time monitoring program is of medium similarity; if so, upload the image captured by the camera during the previous vehicle delivery re-inspection phase to the cloud for manual review; if not, determine that the vehicle is in good condition and meets the rental conditions, and send a vehicle return permission signal to the cloud; Step 14: The program returns to step 1.

2. The automatic collection and detection system for collision status of shared vehicles according to claim 1, characterized in that: The rental car inspection procedure includes: Step 1: To start, the user unlocks the rental car by scanning the QR code on the mobile client, and the controller recognizes the rental car signal; Step 2: Determine whether the vehicle status stored in the body controller is consistent with the phase signal status fed back by the cloud; Step 3: If the two are consistent, the cloud platform sends the vehicle status information uploaded to the cloud at the end of the previous rental via 5G network signals, that is, the car image captured by the camera; Step 4: The vehicle condition information recognition controller reads the vehicle status information sent by the cloud platform and temporarily stores it inside the controller; at the same time, the vehicle condition information recognition controller calls the automatic telescopic device startup program; Step 5: After the automatic retractable device startup program is completed, the vehicle condition photo similarity comparison program is called to perform a similarity comparison on the set of images to determine whether the user's driving process has caused damage to the vehicle; Step 6: Return the result to the main program.

3. The automatic collection and detection system for collision status of shared vehicles according to claim 1, characterized in that: The vehicle delivery re-inspection procedure includes: Step 1: The user sends a vehicle return command to the cloud through the mobile client, and the vehicle condition information recognition controller recognizes the vehicle return signal; Step 2: Determine whether the vehicle status stored in the body controller is consistent with the phase signal status fed back by the cloud; Step 3: If the two are consistent, the cloud platform sends the vehicle status information uploaded to the cloud at the beginning of the rental via 5G network signals, that is, the car image captured by the camera; Step 4: The vehicle condition information recognition controller reads the vehicle status information sent by the cloud platform and temporarily stores it internally; at the same time, the vehicle condition information recognition controller calls the automatic telescopic device startup program; Step 5: After the automatic detection program is completed, the vehicle condition photo similarity comparison program is called to perform a similarity comparison on the set of images to determine whether the user's driving process has caused damage to the vehicle condition; Step 6: Return the result to the main program.

4. The automatic collection and detection system for collision status of shared vehicles according to claim 2 or 3, characterized in that: The vehicle condition photo similarity comparison procedure includes: Step 1: The vehicle condition information recognition controller retrieves the two most recently updated vehicle condition photos stored in it to form a control group for subsequent image comparison; Step 2: Determine the similarity parameters of the group of photos using a comparison program and record them in the controller; Step 3: The comparison program sets a similarity upper limit S 上 and the similarity lower limit S 下 Two values, When the similarity is between 0 and S 下 When the vehicle condition at this time is between 0 and 1, the vehicle condition at this time is defined as a high-difference vehicle condition, and it is determined that the vehicle has been bumped or scratched in the previous stage; When the similarity is between S 下 and S 上 When the vehicle condition is between 0 and 1, the vehicle condition at this time is defined as a medium difference vehicle condition, and further comprehensive analysis of the vehicle condition at this time is required based on the results obtained by the directional sound sensor; When the similarity is greater than S 上 When , the car is defined as undamaged; Step 4: If medium or high difference in vehicle conditions occurs, the image information will be fed back to the cloud via the 5G network to inform the platform personnel of the damage to the vehicle. The cloud platform will then inform the user through the user's mobile client APP or negotiate with the user on compensation matters.

5. The automatic collection and detection system for collision status of shared vehicles according to claim 1, characterized in that: The directional sound sensor real-time monitoring program includes: Step 1: Read the vehicle status information stored in the body controller; Step 2: Determine whether the vehicle status information is consistent with the program stage information; if the judgment result is no, end the program and return the corresponding result; if the judgment result is yes, proceed to the next step; Step 3: receiving signals from four vehicle body directional sound sensors and two chassis directional sound sensors and respectively calling the vehicle body directional sound similarity comparison program and the chassis directional sound sensor comparison program to perform real-time judgment; Step 4: Determine whether the comparison result is abnormal; if so, store the position of the sensor corresponding to the abnormal sound signal and the similarity calculated by the comparison program in the vehicle condition information recognition controller and proceed to the next step; if not, directly proceed to the next step; Step 5: Check whether the program stage information stored in the vehicle condition information recognition controller has changed; if not, continue to step 3; if yes, end the program.

6. The automatic collection and detection system for collision status of shared vehicles according to claim 5, characterized in that: The vehicle body directional sound similarity comparison procedure includes: Step 1: Receive signals from four vehicle body directional sound sensors; Step 2: The sound intensity S of the directional sound sensor signals received in real time at the front, rear, left and right sides of the car n , n = 1, 2, 3, 4 are compared with the sound intensity S0 of the directional sound sensor signal when the vehicle body is bumped in advance, and the corresponding similarity M is obtained. n , n=1,2,3,4, that is Step 3: Compare and get the corresponding similarity M n ,n=1,2,3,4 respectively with M 下 Compare and judge M n >M 下 Is it established? 下 is the similarity threshold for no scratches or bumps, and its specific value is determined by experiments. If the judgment result of the similarity corresponding to a certain sensor is no, then the similarity is judged to be low, and the car in the direction corresponding to the sensor has not experienced any scratches or bumps, and the process continues to step 6. If the judgment result is yes, then the similarity continues to participate in the next step of the process. Step 4: Match the judgment formula M in step 3 n >M 下 The similarity M n , respectively with M 上 Compare and judge M n >M 上 Is it established? 上 is the similarity threshold for the car to only experience scratches, and its specific value is determined by experiments. If the judgment result of the similarity corresponding to a certain sensor is no, then the similarity is judged to be medium, and the car in the direction corresponding to the sensor only experiences scratches. If the judgment result is yes, then the similarity is judged to be high, and the car in the direction corresponding to the sensor experiences both scratches and bumps. Step 5: Return to the position corresponding to the directional sound sensor in step 4; Step 6: Return the corresponding result and the program ends.

7. The automatic collection and detection system for collision status of shared vehicles according to claim 5, characterized in that: The chassis directional sound similarity comparison procedure includes: Step 1: Receive signals from two chassis directional sound sensors; Step 2: The sound intensity S of the directional sound sensor signal received in real time at the front and rear ends of the chassis n , n = 5, 6 are compared with the sound intensity S'0 of the directional sound sensor signal when the chassis is bumped in advance to obtain the corresponding similarity M n , n=5,6, that is Step 3: Compare and get the corresponding similarity M n , n=5,6 respectively with M 安全 Compare and judge M n >M 安全 Is it established? 安全 The threshold for similarity that indicates no serious scratches or bumps on the chassis is determined experimentally. If a sensor's similarity is negative, the similarity is considered safe, indicating no serious scratches or bumps on the chassis at the location corresponding to the sensor. If a directional sound sensor's similarity is positive, the similarity is considered dangerous, indicating serious scratches or bumps on the chassis at the location corresponding to the directional sound sensor. The location of the directional sound sensor is then returned to the cloud. Step 4: Return the corresponding result of step 3 to the cloud. The cloud records the dangerous similarity and the rental user information for subsequent maintenance and accountability. This process ends.

8. The automatic collection and detection system for collision status of shared vehicles according to claim 2 or 3, characterized in that: The automatic telescopic device startup procedure includes: Step 1: Read the straight-line distance between the end of the front automatic telescopic device and the front obstacle measured by the ultrasonic radar, record it as L1, and the straight-line distance between the end of the rear automatic telescopic device and the rear obstacle, record it as L2, and store the data in the vehicle condition information recognition controller; Step 2: Based on the straight-line distances L1 and L2, the controller calculates the telescopic distances D1 and D2 of the telescopic rods in the front and rear automatic telescopic devices according to the following formula; Where n = 1, 2 Determine whether D1 and D2 have reached the minimum distance. If the minimum distance is not reached, feedback is given to the user to adjust the vehicle position. After the user adjusts, the program is triggered again. If the user cannot find a suitable position, the manual photo upload option is selected. If the minimum distance is exceeded, the corresponding control signals are sent to the two telescopic drive motors T1 and T2 according to the D1 and D2 control instructions; Step 3: After receiving the signal, the two telescopic drive motors T1 and T2 drive their respective telescopic rods to reach the telescopic distances D1 and D2 respectively; Step 4: Based on the straight-line distances L1 and L2 and other known quantities, the controller calculates the first rotation angles γ1 and γ2 of the front and rear telescopic cameras according to the following formula, where H is the vertical distance between the spatial position where the camera finally stops moving to capture the vehicle body and the starting point of the telescopic rod of the front automatic telescopic device; Where n = 1, 2 It is sent as a control instruction to the two rotating drive motors T ' 1. T ' 2. Motor T ' 1. T ' 2. Rotate the corresponding angles to drive each camera to a reasonable position, and simultaneously start the front camera, rear camera, left vehicle camera, and right vehicle camera arranged around the vehicle body to take the first photo, and store the collected vehicle condition photos in the controller; Step 5: Based on the straight-line distances L1, L2 and other known quantities, the controller again calculates the second rotation angle γ of the front camera and the rear telescopic camera according to the following formula: ' 1. γ ' 2, where L represents the straight-line distance between the ends of the front and rear telescopic rods and the front and rear obstacles. Where n = 1, 2 It is sent as a control instruction to the two rotating drive motors T ' 1. T ' 2. Motor T ' 1. T ' 2. Rotate the corresponding angles to drive each camera to a reasonable position, and at the same time start the front camera, rear camera, left vehicle camera and right vehicle camera arranged around the vehicle body to take a second photo, and store the collected vehicle condition photos in the controller; Step 6: Control the two rotation drive motors to drive the front camera and the rear camera to rotate in opposite directions γ ' 1. γ ' 2 angles, return to their respective initial positions; Step 7: Control the two telescopic drive motors T1 and T2 to drive the telescopic rod to retract to the initial position, and the program ends; It is also characterized in that the variables in step 2, step 4, and step 5 are defined as follows: α is the fixed angle formed by the telescopic rod of the automatic telescopic device and the horizontal plane, S is the spacing margin reserved for safety; β1 is the angle between the line connecting points B and G and the vertical direction; β2 is the angle between the line connecting points A and G and the vertical direction; β3 is the angle between the line connecting points C and G and the vertical direction, is the effective shooting range angle of the camera in the vertical direction; M is the horizontal distance between points C and G; K is the vertical distance between points C and A; N is the vertical distance between points A and B; among them, point A is the lowest point of the car bumper, point B is the starting point of the telescopic rod of the automatic telescopic device, point C is the top point of the car window, and point G is the spatial position point where the front camera or the rear camera last stops moving to shoot the car body.

Images