A bumper collision intelligent warning method

Abstract

The application discloses a bumper collision intelligent warning method, relates to the technical field of bumper intelligent warning, and saves the captured front bumper image, rear bumper image, front image in front of the front bumper and rear image behind the rear bumper through real-time receiving, and then whether the vehicle bumper is collided is monitored in real time; when the vehicle bumper is collided, the bumper position collided and the license plate number of the vehicle causing the collision are obtained according to the front image in front of the front bumper and the rear image behind the rear bumper; then, the corresponding collision bumper image is obtained based on the bumper position collided; the bumper damage severity is analyzed based on the collision bumper image; finally, the collision alarm information is sent to the intelligent mobile terminal of the vehicle owner; the remote safety monitoring of the vehicle is realized, and the legal rights and interests of the vehicle owner are maintained.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of bumper collision warning, and specifically relates to an intelligent bumper collision warning method. BACKGROUND

[0002] With the increase of vehicle use, vehicle collision events also occur from time to time. However, the current vehicle safety monitoring system is mostly limited to monitoring when the vehicle owner is in the vehicle, and cannot timely alarm and record the collision situation after the vehicle owner leaves the vehicle. This makes the vehicle owner unable to effectively maintain his own legal rights and interests, and the remote safety monitoring of the vehicle has limitations.

[0003] The traditional vehicle safety monitoring system is mainly based on sensors and alarm devices inside the vehicle, but these systems cannot effectively monitor the vehicle collision situation after the vehicle owner leaves the vehicle. After the vehicle collision, the vehicle owner often cannot timely learn about the collision event, and cannot provide records and evidence at the time of collision, resulting in difficulties in maintaining his own rights and interests.

[0004] Therefore, the application provides an intelligent bumper collision warning method. SUMMARY

[0005] The application aims to at least solve one of the technical problems existing in the prior art. To this end, the application provides an intelligent bumper collision warning method, which realizes remote safety monitoring of the vehicle and maintains the legal rights and interests of the vehicle owner.

[0006] To achieve the above-mentioned purpose, according to the embodiment of the first aspect of the application, an intelligent bumper collision warning method is provided, comprising the following steps:

[0007] The cloud server end receives and saves the front bumper image and the rear bumper image captured by the first image capturing device and the second image capturing device installed on the vehicle, the front bumper front image captured by the third image capturing device, and the rear bumper rear image captured by the fourth image capturing device in real time;

[0008] The cloud server end monitors whether the vehicle bumper is collided in real time, and if the vehicle bumper is collided:

[0009] The cloud server end obtains the position of the collided bumper and the license plate number of the vehicle involved in the collision according to the front bumper front image and the rear bumper rear image;

[0010] The cloud server end obtains the corresponding collision bumper image based on the position of the collided bumper, and analyzes the damage severity of the bumper based on the collision bumper image;

[0011] The cloud server end sends the collision alarm information to the intelligent mobile terminal of the vehicle owner.

[0012] The mounting positions of the first image capturing device and the second image capturing device on the vehicle are positions from which the front bumper and the rear bumper panoramas can be captured, respectively; the specific mounting positions of the first image capturing device and the second image capturing device are determined according to different shapes and layouts of different vehicles;

[0013] The third image capturing device is mounted on the front of the vehicle, and the image capturing mechanism thereof is directed forward;

[0014] The fourth image capturing device is mounted on the rear of the vehicle, and the image capturing mechanism thereof is directed rearward;

[0015] The specific mounting positions of the third image capturing device and the fourth image capturing device are determined according to different shapes and layouts of different vehicles, so as to ensure that the license plate numbers of vehicles in front of and behind the vehicle can be captured;

[0016] The data of the front bumper image, the rear bumper image, the front bumper front image, and the rear bumper rear image are saved in the storage device of the cloud server;

[0017] The cloud server monitors in real time whether the vehicle bumper is collided in the following manner:

[0018] The vehicle speed sensor, the front vibration sensor, and the rear vibration sensor mounted on the vehicle send the vehicle speed, the front vibration value, and the rear vibration value to the cloud server in real time, respectively; the front vibration sensor is mounted on the front bumper, and the rear vibration sensor is mounted on the rear bumper;

[0019] The cloud server determines that the vehicle is collided when the vehicle speed is less than a preset vehicle speed threshold value and the vibration value is greater than a preset vibration value threshold value; otherwise, the cloud server determines that the vehicle is not collided;

[0020] The license plate number of the collided bumper position and the vehicle involved in the collision is obtained in the following manner:

[0021] The cloud server compares the front vibration value and the rear vibration value at the current time,

[0022] If the front vibration value is greater, it is determined that the front bumper is collided, and the front bumper front image saved in the storage device is retrieved as the collision detection image;

[0023] If the rear vibration value is greater, it is determined that the rear bumper is collided, and the rear bumper rear image saved in the storage device is retrieved as the collision detection image;

[0024] An image recognition algorithm is used to recognize the license plate of the vehicle appearing most recently in the collision detection image, and an OCR technology is used to recognize the license plate number from the license plate;

[0025] The corresponding collision bumper image is obtained in the following manner:

[0026] If the collision detection image is the front bumper front image, the bumper image saved in the storage device is taken as the collision bumper image;

[0027] If the collision detection image is the rear bumper rear image, the bumper image saved in the storage device is taken as the collision bumper image;

[0028] The way of calculating the bumper damage severity based on the bumper image is:

[0029] A plurality of damage levels are set in advance, the number of damage levels is marked as N, and the number of damage levels is marked as n;

[0030] A preset comparison time t is set, and the cloud server end calls the collision bumper image at the current time and the collision bumper image before the collision from the storage device; the collision bumper image before the collision is the collision bumper image at the comparison time t before the bumper of the vehicle is judged to be collided;

[0031] The collision bumper image at the current time is marked as the first collision bumper image, and the collision bumper image before the collision is marked as the second collision bumper image;

[0032] The target recognition algorithm is used on the first collision bumper image to identify whether the bumper at the corresponding position is contained; if the bumper is not identified, it is judged that the bumper is knocked off or has been severely deformed, and the bumper damage severity is set to the Nth damage level;

[0033] If the bumper is identified, the damage coefficient of the bumper at the corresponding position is calculated;

[0034] The way of calculating the damage coefficient of the bumper at the corresponding position is:

[0035] The target recognition algorithm is used to identify the bumper in the first collision bumper image and the second collision bumper image respectively, and the image of the bumper part in each is intercepted; the image of the bumper part in the first collision bumper image is marked as the first analysis image, and the image of the bumper part in the second collision bumper image is marked as the second analysis image;

[0036] The target recognition algorithm is used to identify all the cracks in the first analysis image, and the number of cracks is marked as L1, the number of cracks is marked as k1, and the first crack damage degree Sk1 of the k1th crack is calculated, wherein the first crack damage degree Sk1 is the total number of pixels occupied by the k1th crack in the first analysis image;

[0037] The first crack total damage degree SL1 of the first analysis image is calculated; wherein the calculation formula of the first crack total damage degree SL1 is SL1 =∑ l1Sk1;

[0038] using a target recognition algorithm, all cracks in the second analysis image are recognized, and the number of the cracks is marked as L2, the number of the cracks is marked as k2, and the second crack damage degree Sk2 of the k2th crack is calculated, wherein the second crack damage degree Sk2 is the total number of pixel points occupied by the k2th crack in the second analysis image;

[0039] The second crack total damage degree SL2 of the second analysis image is calculated; wherein the calculation formula of the second crack total damage degree SL2 is SL2 = ∑ l2 Sk2;

[0040] using an edge detection algorithm, all scratches in the first analysis image are detected, and the area of each scratch is calculated; wherein the area of each scratch is the total number of pixel points of the region in the first analysis image; the total area of all scratches in the first analysis image is calculated as the first scratch total damage degree; and the first scratch total damage degree is marked as SG1;

[0041] using an edge detection algorithm, all scratches in the second analysis image are detected, and the area of each scratch is calculated; wherein the area of each scratch is the total number of pixel points of the region in the second analysis image; the total area of all scratches in the second analysis image is calculated as the second scratch total damage degree; and the second scratch total damage degree is marked as SG2;

[0042] The damage coefficient S of the bumper is obtained, and the calculation formula of the damage coefficient S is S = a1*(SL1-SL2) + a2*(SG1-SG2); wherein a1 and a2 are respectively preset proportion coefficients;

[0043] The damage coefficient S is obtained based on the damage coefficient S to obtain the damage severity of the bumper;

[0044] The way to obtain the damage severity of the bumper based on the damage coefficient S is:

[0045] The cloud server side sets a damage coefficient lower limit value for each damage severity level in advance, and the damage coefficient lower limit value of the nth damage severity level is marked as Dn, and the damage coefficient range of the nth damage severity level is between Dn and D(n+1);

[0046] The damage coefficient range corresponding to the damage severity level of the bumper is determined by searching the damage coefficient range corresponding to the value of the damage coefficient S;

[0047] The way for the cloud server side to send the collision alarm information to the intelligent mobile terminal of the vehicle owner is:

[0048] The cloud server end is connected with the intelligent mobile terminal of the vehicle owner in real time through a wireless network, and after determining the damage severity of the bumper, sends a collision alarm information to the intelligent mobile terminal of the vehicle owner through the wireless network; the collision alarm information includes the position of the collided bumper, the license plate number of the vehicle causing the collision, the image of the collided bumper and the damage severity of the bumper.

[0049] Compared with the prior art, the beneficial effects of the present application are:

[0050] The present application captures the images outside the vehicle and the images of the bumpers in real time by installing two image capturing devices at the front bumper and the rear bumper respectively, judges whether the vehicle is collided by using the vehicle speed sensor and the vibration sensor, obtains the position of the collided bumper and the license plate number of the vehicle causing the collision when the vehicle is collided, further obtains the corresponding image of the collided bumper, analyzes the damage severity of the bumper based on the image of the collided bumper, and finally sends the collision alarm information to the client of the vehicle owner; ensures that the collision record is kept in time and the vehicle owner is alarmed in the case that the vehicle is collided after the vehicle owner leaves the vehicle, realizes the remote safety monitoring of the vehicle, and maintains the legal rights and interests of the vehicle owner. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 The flow chart of the bumper collision intelligent alarm method in embodiment 1 of the present application. DETAILED DESCRIPTION

[0052] The technical solutions of the present application will be described clearly and completely in combination with embodiments. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0053] Embodiment 1

[0054] As shown in Figure 1 A bumper collision intelligent alarm method, comprising the following steps:

[0055] Step 1: The cloud server end receives and saves the front bumper image and the rear bumper image captured by the first image capturing device and the second image capturing device installed on the vehicle, the image in front of the front bumper captured by the third image capturing device and the image behind the rear bumper captured by the fourth image capturing device in real time;

[0056] Step 2: The cloud server end monitors whether the vehicle bumper is collided in real time, if the vehicle bumper is collided, go to step 3; if the vehicle bumper is not collided, continue to monitor;

[0057] Step three: the cloud server obtains the position of the collided bumper and the license plate number of the vehicle based on the front bumper front image and the rear bumper rear image;

[0058] Step four: the cloud server obtains the corresponding collision bumper image based on the position of the collided bumper, and analyzes the bumper damage severity based on the collision bumper image;

[0059] Step five: the cloud server sends a collision alarm information to the intelligent mobile terminal of the vehicle owner;

[0060] In a preferred embodiment, the installation positions of the first image capturing device and the second image capturing device on the vehicle are positions capable of capturing panoramic images of the front bumper and the rear bumper respectively; the specific installation positions of the first image capturing device and the second image capturing device are determined according to the different shapes and layouts of different vehicles;

[0061] In the present application, the front and rear correspondences are that the direction from the tail to the head is the front, and the direction from the head to the tail is the rear;

[0062] The third image capturing device is installed at the head, and the direction of its image capturing mechanism points to the front of the vehicle;

[0063] The fourth image capturing device is installed at the tail, and the direction of its image capturing mechanism points to the rear of the vehicle;

[0064] The specific installation positions of the third image capturing device and the fourth image capturing device are determined according to the different shapes and layouts of different vehicles, so as to ensure that the license plate numbers of the vehicles in front and rear of the vehicle can be captured;

[0065] The data of the front bumper image, the rear bumper image, the front bumper front image and the rear bumper rear image are saved in the storage device of the cloud server; preferably, a saving time is set for the saving time length of all images, and the images exceeding the saving time are automatically deleted;

[0066] Further, the cloud server end monitors whether the vehicle bumper is collided in real time in the following way:

[0067] The vehicle speed sensor, the front vibration sensor and the rear vibration sensor installed on the vehicle send the vehicle speed, the front vibration value and the rear vibration value to the cloud server end in real time; the front vibration sensor is installed on the front bumper, and the rear vibration sensor is installed on the rear bumper, which can more effectively determine the collision of the bumper;

[0068] When the vehicle speed is less than a preset vehicle speed threshold value and the vibration value is greater than a preset vibration value threshold value, the cloud server end determines that the vehicle is collided; otherwise, it is determined that the vehicle is not collided;

[0069] The way to obtain the position of the collided bumper and the license plate number of the vehicle causing the accident is:

[0070] The cloud server compares the size of the front vibration value and the rear vibration value at the current time,

[0071] If the front vibration value is larger, it is determined that the front bumper is collided, and the image in front of the front bumper saved in the storage device is retrieved as the collision detection image;

[0072] If the rear vibration value is larger, it is determined that the rear bumper is collided, and the image behind the rear bumper saved in the storage device is retrieved as the collision detection image;

[0073] The license plate of the vehicle appearing most recently in the collision detection image is identified using an image recognition algorithm, and the license plate number is identified from the license plate using OCR technology;

[0074] The way to obtain the corresponding collision bumper image is:

[0075] If the collision detection image is the image in front of the front bumper, the front bumper image saved in the storage device is taken as the collision bumper image;

[0076] If the collision detection image is the image behind the rear bumper, the rear bumper image saved in the storage device is taken as the collision bumper image;

[0077] The way to calculate the damage severity of the bumper based on the bumper image is:

[0078] A number of damage levels are set in advance, the number of damage levels is marked as N, and the number of damage levels is marked as n;

[0079] A preset comparison time t is set, and the cloud server side retrieves the collision bumper image at the current time and the collision bumper image before the collision from the storage device; The collision bumper image before the collision is the collision bumper image at the comparison time t before the bumper of the vehicle is determined to be collided;

[0080] The collision bumper image at the current time is marked as the first collision bumper image, and the collision bumper image before the collision is marked as the second collision bumper image;

[0081] The target recognition algorithm is used on the first collision bumper image to identify whether the bumper at the corresponding position is included; If the bumper is not identified, it is determined that the bumper is knocked off or has been severely deformed, and the damage severity of the bumper is set to the Nth damage level;

[0082] It can be understood that, due to the difference in shapes of the front and rear bumpers, the target identified by the target identification algorithm is also different, and the corresponding position is the front bumper or the rear bumper corresponding to the collision bumper image; the target identification algorithm belongs to the commonly used technology in the art, and by collecting a plurality of bumper images as training data of a CNN neural network model, the CNN neural network model is trained, and a CNN neural network model for identifying bumpers is obtained;

[0083] If the bumper is identified, the damage coefficient of the bumper at the corresponding position is calculated;

[0084] In a preferred embodiment, the damage coefficient of the bumper at the corresponding position is calculated in the following manner:

[0085] The target identification algorithm is used to identify the bumpers in the first and second collision bumper images respectively, and the images of the bumper parts are intercepted respectively; the image of the bumper part of the first collision bumper image is marked as a first analysis image, and the image of the bumper part of the second collision bumper image is marked as a second analysis image;

[0086] Using the target identification algorithm, all the cracks in the first analysis image are identified, and the number of the cracks is marked as L1, and the number of the cracks is marked as k1, and the first crack damage degree Sk1 of the k1th crack is calculated, wherein the first crack damage degree Sk1 is the total number of pixels occupied by the k1th crack in the first analysis image;

[0087] The first crack total damage degree SL1 of the first analysis image is calculated; wherein the calculation formula of the first crack total damage degree SL1 is SL1 =∑ l1 Sk1;

[0088] Using the target identification algorithm, all the cracks in the second analysis image are identified, and the number of the cracks is marked as L2, and the number of the cracks is marked as k2, and the second crack damage degree Sk2 of the k2th crack is calculated, wherein the second crack damage degree Sk2 is the total number of pixels occupied by the k2th crack in the second analysis image;

[0089] The second crack total damage degree SL2 of the second analysis image is calculated; wherein the calculation formula of the second crack total damage degree SL2 is SL2 =∑ l2 Sk2;

[0090] all scratch areas in the first analysis image are detected using an edge detection algorithm, and the area of each scratch area is calculated; wherein the area of each scratch area is the total number of pixel points of the area in the first analysis image; the total area of all scratch areas in the first analysis image is calculated as the first total scratch damage degree; and the first total scratch damage degree is marked as SG1; it can be understood that, after the bumper is scratched, the paint on the surface will be scratched, resulting in a color difference between the scratch area and other areas, so an edge detection algorithm based on pixel point changes can be used to detect the scratch area;

[0091] all scratch areas in the second analysis image are detected using an edge detection algorithm, and the area of each scratch area is calculated; wherein the area of each scratch area is the total number of pixel points of the area in the second analysis image; the total area of all scratch areas in the second analysis image is calculated as the second total scratch damage degree; and the second total scratch damage degree is marked as SG2;

[0092] the damage coefficient S of the bumper is obtained, and the calculation formula of the damage coefficient S is S = a1*(SL1-SL2) + a2*(SG1-SG2); it can be understood that the damage coefficient S is the loss caused by the bumper in this collision; wherein a1 and a2 are respectively preset proportion coefficients;

[0093] the damage coefficient S is obtained based on the damage coefficient S to obtain the bumper damage severity;

[0094] the way to obtain the bumper damage severity based on the damage coefficient S is:

[0095] the cloud server end pre-sets a damage coefficient lower limit value for each damage level, and marks the damage coefficient lower limit value of the nth damage level as Dn, so the damage coefficient range of the nth damage level is between Dn and D(n+1);

[0096] the damage coefficient range corresponding to the value of the damage coefficient S is found, and the bumper damage severity is determined as the damage level corresponding to the damage coefficient range;

[0097] the way for the cloud server end to send the collision alarm information to the smart mobile terminal of the vehicle owner is:

[0098] the cloud server end and the smart mobile terminal of the vehicle owner are connected in real time in a wireless network mode, and after the bumper damage severity is determined, the collision alarm information is sent to the smart mobile terminal of the vehicle owner through the wireless network; the collision alarm information includes the position of the collided bumper, the license plate number of the vehicle involved in the accident, the image of the collided bumper, and the bumper damage severity.

[0099] The above examples are only used to illustrate the technical method of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present application can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present application.

Claims

1. A bumper collision intelligent warning method, characterized by, The method comprises the following steps: The cloud server end receives and saves the front bumper image and the rear bumper image captured by the first image capturing device and the second image capturing device installed on the vehicle, the front bumper front image captured by the third image capturing device, and the rear bumper rear image captured by the fourth image capturing device in real time; The cloud server end monitors whether the vehicle bumper is collided in real time, and when the vehicle bumper is collided: The cloud server end obtains the position of the collided bumper and the license plate number of the vehicle involved in the collision according to the front bumper front image and the rear bumper rear image; The cloud server end obtains the corresponding collision bumper image based on the position of the collided bumper, and analyzes the damage severity of the bumper based on the collision bumper image; The cloud server end sends a collision alarm information to the intelligent mobile terminal of the vehicle owner; The analysis of the damage severity of the bumper based on the collision bumper image comprises calculating the damage coefficient of the bumper based on the collision bumper image to determine the damage severity of the bumper, wherein the calculation method of the damage coefficient of the bumper is as follows: The target recognition algorithm is used to identify the bumper in the first collision bumper image and the second collision bumper image respectively, and the image of the bumper part in each image is intercepted; the image of the bumper part in the first collision bumper image is marked as a first analysis image, and the image of the bumper part in the second collision bumper image is marked as a second analysis image; The target recognition algorithm is used to identify all the cracks in the first analysis image, and the number of the cracks is marked as L1, the number of the cracks is marked as k1, and the first crack damage degree Sk1 of the k1th crack is calculated, wherein the first crack damage degree Sk1 is the total number of the pixel points occupied by the k1th crack in the first analysis image; a first total crack damage SL1 of the first analysis image is calculated; wherein, a calculation formula of the first total crack damage SL1 is ; The target recognition algorithm is used to identify all the cracks in the second analysis image, and the number of the cracks is marked as L2, the number of the cracks is marked as k2, and the second crack damage degree Sk2 of the k2th crack is calculated, wherein the second crack damage degree Sk2 is the total number of the pixel points occupied by the k2th crack in the second analysis image; calculating a second total crack damage SL2 of the second analysis image; wherein the formula of the second total crack damage SL2 is ; The edge detection algorithm is used to detect all the scratch areas in the first analysis image, and the area of each scratch area is calculated; wherein the area of each scratch area is the total number of the pixel points of the area in the first analysis image; the total area of all the scratch areas in the first analysis image is calculated as a first total scratch damage degree; and the first total scratch damage degree is marked as SG1; The edge detection algorithm is used to detect all the scratch areas in the second analysis image, and the area of each scratch area is calculated; wherein the area of each scratch area is the total number of the pixel points of the area in the second analysis image; the total area of all the scratch areas in the second analysis image is calculated as a second total scratch damage degree; and the second total scratch damage degree is marked as SG2; The damage coefficient S of the bumper is calculated as S=a1*(SL1-SL2)+a2*(SG1-SG2); wherein a1 and a2 are preset proportion coefficients.

2. The bumper collision intelligent warning method according to claim 1, wherein, The mounting positions of the first image capturing device and the second image capturing device on the vehicle are positions from which the front bumper and the rear bumper panoramas can be captured, respectively; The third image capturing device is mounted on the front of the vehicle, and the direction of its image capturing mechanism points to the front of the vehicle; The fourth image capturing device is mounted on the rear of the vehicle, and the direction of its image capturing mechanism points to the rear of the vehicle; The data of the front bumper image, the rear bumper image, the front bumper front image, and the rear bumper rear image are saved in the storage device of the cloud server.

3. The bumper collision intelligent warning method of claim 2, wherein, The cloud server monitors whether the vehicle bumper is collided in real time in the following manner: The vehicle speed sensor, the front vibration sensor, and the rear vibration sensor installed on the vehicle send the vehicle speed, the front vibration value, and the rear vibration value to the cloud server in real time, respectively; the front vibration sensor is installed on the front bumper, and the rear vibration sensor is installed on the rear bumper; The cloud server judges that the vehicle is collided when the vehicle speed is less than a preset vehicle speed threshold value, and the vibration value is greater than a preset vibration value threshold value.

4. The bumper collision intelligent warning method according to claim 3, wherein, The manner of obtaining the position of the collided bumper and the license plate number of the vehicle involved in the accident is as follows: The cloud server compares the front vibration value and the rear vibration value at the current time, If the front vibration value is greater, it is determined that the front bumper is collided, and the front bumper front image saved in the storage device is called as the collision detection image; If the rear vibration value is greater, it is determined that the rear bumper is collided, and the rear bumper rear image saved in the storage device is called as the collision detection image; The image recognition algorithm is used to identify the license plate of the vehicle appearing most recently from the collision detection image, and the OCR technology is used to identify the license plate number from the license plate.

5. The bumper crash smart warning method of claim 4, wherein, The manner of obtaining the corresponding collision bumper image is as follows: If the collision detection image is the front bumper front image, the front bumper image saved in the storage device is taken as the collision bumper image; If the collision detection image is the rear bumper rear image, the rear bumper image saved in the storage device is taken as the collision bumper image.

6. The bumper crash smart warning method of claim 5, wherein, The manner of calculating the damage severity of the bumper based on the bumper image is as follows: A plurality of damage levels are set in advance, the number of damage levels is marked as N, and the number of damage levels is marked as n; A preset comparison time t is set, and the cloud server calls the collision bumper image at the current time and the collision bumper image before the collision from the storage device; the collision bumper image before the collision is the collision bumper image at the comparison time t before it is judged that the vehicle bumper is collided; The collision bumper image at the current time is marked as the first collision bumper image, and the collision bumper image before the collision is marked as the second collision bumper image; The target recognition algorithm is used to identify whether the bumper at the corresponding position is contained in the first collision bumper image; If the bumper is not identified, it is judged that the bumper is knocked off or has been severely deformed, and the damage severity of the bumper is set to the Nth damage level; If the bumper is identified, the damage coefficient of the bumper at the corresponding position is calculated; The damage severity of the bumper is obtained based on the damage coefficient.

7. The bumper crash smart warning method of claim 6, wherein, The manner of obtaining the damage severity of the bumper based on the damage coefficient S is as follows: The cloud server end sets a damage coefficient lower limit value for each damage level in advance, and marks the damage coefficient lower limit value of the nth damage level as Dn, so that the damage coefficient range of the nth damage level is between Dn and D(n+1); The damage coefficient range corresponding to the value of the damage coefficient S is found to determine the bumper damage severity as the damage level corresponding to the damage coefficient range.

8. The bumper crash smart warning method of claim 7, wherein, The cloud server end sends the bumper collision warning information to the smart mobile terminal of the vehicle owner in the following manner: The cloud server end is connected with the smart mobile terminal of the vehicle owner in real time through a wireless network, and sends the bumper collision warning information to the smart mobile terminal of the vehicle owner through the wireless network after determining the bumper damage severity. The bumper collision warning information includes the bumper position, the license plate number of the vehicle involved in the accident, the bumper collision image, and the bumper damage severity.

Images