Anti-cheating method and anti-cheating system for transport vehicle

Abstract

The invention relates to an anti-cheating method and an anti-cheating system for a transport vehicle, and the method comprises the following steps: S1, when the vehicle enters a weighing platform region, a camera detects the license plate of a vehicle head and starts to shoot, and the shot content comprises the license plate of the vehicle head, the license plate of a vehicle body and the license plate of a vehicle tail; s2, a license plate number and a complete vehicle side panoramic image are obtained according to the shot image, and the license plate number and the vehicle side panoramic image are stored in an associated mode; s3, establishing a data set capable of being trained by a deep learning algorithm according to the vehicle side panoramic image, forming a feature extraction and feature matching model through training, and forming a feature matching algorithm; s4, a post-processing logic program of a feature matching result is formed based on a feature matching algorithm, the post-processing logic program comprises normalization processing and similarity score calculation, normalization processing is carried out on two vehicle side panoramic images obtained when the same vehicle enters and leaves the factory area, and a similarity score is calculated; and S5, comparing the similarity score with a set threshold value, if the similarity score is greater than the similarity threshold value, determining that the vehicle is the same vehicle, and if the similarity score is less than the similarity threshold value, determining that the vehicle is not the same compartment, and carrying out alarm prompt. According to the method, whether the carriages of the incoming and outgoing vehicles are replaced or not is judged according to the similarity of the side views of the vehicles, the weighing cheating behavior of replacing the carriages is prevented, and the metering loss of enterprises caused by the weighing cheating behavior is reduced.

Description

Technical Field

[0001] The present invention relates to the field of monitoring technology, and more specifically, to a transportation vehicle anti-cheating method and an anti-cheating system. Background Art

[0002] In steel enterprises, there are many vehicles transporting materials or products in and out of the factory, and the types of vehicles are diverse. The vehicles transporting materials or products need to be weighed on the scale. At present, the weighing method of the scale is generally automatic weighing. There is no full-time personnel to monitor the weighing process of the vehicle. It is impossible to determine whether the vehicle entering and leaving the vehicle has replaced the carriage, and it is impossible to prevent the occurrence of cheating in weighing by replacing the carriage. Therefore, it is necessary to propose an intelligent and automated means to detect whether there is cheating in the weighing process of the vehicle. Summary of the invention

[0003] In view of the shortcomings of the prior art, the purpose of the present invention is to provide an anti-cheating method and anti-cheating system for transport vehicles, which can judge whether the carriages of vehicles entering and leaving the factory have been replaced based on the similarity of the vehicle side views, prevent the occurrence of weighing cheating by replacing the carriages, and reduce the measurement losses suffered by the enterprise.

[0004] The above technical objectives of the present invention are achieved through the following technical solutions:

[0005] A method for preventing cheating in transport vehicles is provided, which uses machine vision to compare the consistency of vehicles and determine whether the carriages of the vehicles have been replaced. The specific steps are as follows:

[0006] S1. A camera is set in front of the weighing platform. When a vehicle enters the weighing platform area, the camera in front of the vehicle detects the license plate of the front of the vehicle and starts shooting. The shooting content includes the license plate of the front, body and rear of the vehicle. When the license plate of the rear of the vehicle is captured, the shooting stops;

[0007] S2, performing image recognition based on the license plate images of the front and rear license plates to obtain the license plate number, performing image stitching based on the image of the vehicle body to obtain a complete vehicle side panoramic image, and storing the license plate number in association with the vehicle side panoramic image;

[0008] S3. Based on the vehicle side panoramic image formed by image acquisition, a data set for deep learning algorithm training is established, and a feature extraction and feature matching model is formed after training, and a feature matching algorithm is formed;

[0009] S4, deploying the trained feature extraction and feature matching model to the backend server, and forming a post-processing logic program of the feature matching result based on the feature matching algorithm, wherein the post-processing logic program includes normalization processing and similarity score calculation, and performing normalization processing and calculating the similarity score on two vehicle side panoramic images obtained when the same vehicle enters and leaves the factory area;

[0010] S5. Based on the post-processing logic program, a similarity score is obtained, and the similarity score is compared with a set threshold. If the similarity score is greater than the similarity threshold, the vehicles are judged to be the same vehicle; if the similarity score is less than the similarity threshold, the vehicles are judged to be different vehicles, and an alarm is issued.

[0011] In one embodiment, in step S1, the camera is a three-eye camera, which respectively captures the front, rear and body of the vehicle, or the number of the cameras is three, which are responsible for capturing the front, rear and body of the vehicle respectively.

[0012] In one of the embodiments, a marking spraying device is further provided outside the weighing platform, and the marking spraying device is on the same side as the camera. The marking spraying device sprays a detection mark on the side of the vehicle compartment, and the detection mark is photographed by the camera when the vehicle enters the weighing platform.

[0013] In one of the embodiments, in step S1, the vehicle stops in front of the weighing platform, the marking spraying device approaches the side of the vehicle and sprays paint at any position on the side of the vehicle according to the image determined by the back-end server to form a detection mark.

[0014] In one embodiment, the detection mark is a plurality of parallel transverse stripes arranged from top to bottom. On the left side of the detection mark, the ends of the plurality of transverse stripes are flush, and on the right side of the detection mark, there is a length difference between the ends of adjacent transverse stripes. The plurality of transverse stripes form a plurality of length differences, and the plurality of length differences are different.

[0015] In one of the embodiments, in step S4, all length differences in the same detection mark are processed to eliminate the influence of the shooting angle.

[0016] In one embodiment, the method of performing data processing on all length differences in the same detection mark is: calculating the ratio of all length differences, or measuring and calculating the length differences after processing the panoramic image of the vehicle side.

[0017] In one embodiment, the sprayed pigment is a self-degradable pigment.

[0018] A transport vehicle anti-cheating system comprises a control module, a camera and a marking spraying device, wherein the marking spraying device comprises a manipulator, a paint tank, a pump and a nozzle, the camera is arranged outside the vehicle entry end of a weighing platform, and the marking spraying device is arranged in front of the camera along the vehicle entry direction, the vehicle first passes through the camera and then the marking spraying device and then enters the weighing platform, the camera, the manipulator and the pump are all connected to the control module, the pump is respectively connected to the paint tank and the nozzle through pipelines, and the pump is arranged on the manipulator.

[0019] In summary, the present invention has the following beneficial effects:

[0020] The present invention determines whether the carriages of vehicles entering or leaving the factory have been replaced according to the similarity of the vehicle side views, thereby preventing the occurrence of cheating in weighing the replaced carriages and reducing the measurement losses suffered by the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a schematic diagram of step S1 of the present invention;

[0022] Figure 2 is a logical schematic diagram of steps S1 and S2 of the present invention;

[0023] Figure 3 It is an overall logic schematic diagram of the present invention. DETAILED DESCRIPTION

[0024] The present invention is described in detail below in conjunction with the accompanying drawings and embodiments.

[0025] It is worth noting that the directional words such as "upper" and "lower" involved in this article are all relative to the viewing angle of the drawings and are only for the convenience of description and cannot be understood as limitations on the technical solution.

[0026] like Figure 1-3 The present invention proposes a method for preventing cheating in transport vehicles, which uses machine vision to compare the consistency of vehicles and determine whether the vehicle compartment has been replaced. The specific steps are as follows:

[0027] S1. A camera is set in front of the weighing platform. When a vehicle enters the weighing platform area, the camera in front of the vehicle detects the license plate of the front of the vehicle and starts shooting. The shooting content includes the license plate of the front, body and rear of the vehicle. When the license plate of the rear of the vehicle is captured, the shooting stops;

[0028] S2, performing image recognition based on the license plate images of the front and rear license plates to obtain the license plate number, performing image stitching based on the image of the vehicle body to obtain a complete vehicle side panoramic image, and storing the license plate number in association with the vehicle side panoramic image;

[0029] S3. Based on the vehicle side panoramic image formed by image acquisition, a data set for deep learning algorithm training is established, and a feature extraction and feature matching model is formed after training, and a feature matching algorithm is formed;

[0030] S4, deploying the trained feature extraction and feature matching model to the backend server, and forming a post-processing logic program of the feature matching result based on the feature matching algorithm, wherein the post-processing logic program includes normalization processing and similarity score calculation, and performing normalization processing and calculating the similarity score on two vehicle side panoramic images obtained when the same vehicle enters and leaves the factory area;

[0031] S5. Based on the post-processing logic program, a similarity score is obtained, and the similarity score is compared with a set threshold. If the similarity score is greater than the similarity threshold, the vehicles are judged to be the same vehicle. If the similarity score is less than the similarity threshold, the vehicles are judged to be different vehicles. An alarm is issued in the business system and handled by the management personnel.

[0032] The present invention determines whether the vehicle has cheated by automatically photographing and identifying whether the compartments of the same vehicle entering and leaving the factory are consistent, which effectively solves the hidden danger of vehicle cheating, improves vehicle identification efficiency without arranging special personnel for monitoring, and greatly improves work efficiency.

[0033] Generally speaking, vehicles enter or leave the factory through different weighing platforms. According to the direction of the vehicle, a camera is set in front of the vehicle entrance of the weighing platform, so that the vehicle can be filmed before entering the weighing platform. It is easy to understand that when vehicles enter and leave the factory through the same weighing platform, cameras are set in front and behind the same weighing platform. Regardless of whether the vehicle enters or leaves the weighing platform, the corresponding camera can be started to film.

[0034] In step S1, the camera is a three-eye camera, which respectively shoots the front, rear and body of the vehicle, or the number of the cameras is three, which are responsible for shooting the front, rear and body of the vehicle, namely, the front camera, the body camera and the rear camera.

[0035] Furthermore, a marking spraying device is provided outside the weighing platform, and the marking spraying device is on the same side as the camera. The marking spraying device sprays a detection mark on the side of the vehicle compartment. When the vehicle enters the weighing platform, the detection mark is photographed by the camera, and is used as one of the auxiliary identification features of the vehicle compartment in the subsequent steps S3 and S4 to facilitate identification. The specific steps are as follows:

[0036] The vehicle stops in front of the weighing platform, and the marking spraying device approaches the side of the car and sprays paint at any position on the side of the car according to the image determined by the back-end server, forming a detection mark, which is used as one of the characteristics of the car later. The sprayed paint is a self-degradable paint, which disappears automatically after a period of time.

[0037] The purpose of the detection mark is to artificially form a feature on the side of the car. During the identification process, the detection mark can serve as a very obvious feature, which facilitates identification and processing.

[0038] Furthermore, the detection mark is a plurality of parallel transverse stripes arranged from top to bottom. On the left side of the detection mark, the ends of the plurality of transverse stripes are flush, and on the right side of the detection mark, there is a length difference between the ends of adjacent transverse stripes. The plurality of transverse stripes form a plurality of length differences, and the plurality of length differences are different.

[0039] It is easy to understand that for multiple detection marks, a unique identity mark can be formed based on the difference in length difference. Generally speaking, the length difference is small, and all the length differences in the same detection mark cannot be effectively distinguished by the naked eye. After being photographed by a camera, all the length differences in a detection mark can be obtained through software analysis and measurement. By comparing whether the detection mark has changed when entering and leaving the factory, it is possible to quickly determine whether there is any change in the car body.

[0040] It should be noted that in the images taken when entering and leaving the factory, the absolute length of the detection mark itself may be different. This is because the shooting angle is different. All length differences in the same detection mark can be processed to eliminate the influence of the shooting angle, such as calculating the ratio of all length differences, or measuring and calculating the length difference after processing the image.

[0041] Feature extraction and feature matching model training is based on deep learning algorithms such as superpoint and supergle for feature extraction and feature matching, and the image matching accuracy of the trained model in the car scene is verified.

[0042] After the development of each module is completed, the developers will work with the operators to test run the system, collect image data that has not been accurately identified, and iterate the algorithm model. During the test run, any problems will be discovered and promptly addressed and improved, and the system will eventually be officially launched to achieve fully automatic and intelligent recognition.

[0043] The present invention proposes a transport vehicle anti-cheating system, including a control module, a camera and a marking spraying device, the marking spraying device including a manipulator, a paint tank, a pump and a nozzle, the camera is arranged outside the vehicle entry end of the weighing platform, along the vehicle entry direction, the marking spraying device is arranged in front of the camera, the vehicle first passes through the camera and then the marking spraying device and then enters the weighing platform, the camera, the manipulator and the pump are all connected to the control module, the pump is respectively connected to the paint tank and the nozzle through pipelines, and the pump is arranged on the manipulator. Before the vehicle enters the factory area, the side of the carriage is first sprayed with a detection mark by the spraying module, and the subsequent carriage is photographed by the camera module to form an image, which is convenient for the control system to identify the vehicle carriage.

[0044] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A method for preventing cheating in transport vehicles, characterized in that: Here are the steps: S1. A camera is set in front of the weighing platform. When a vehicle enters the weighing platform area, the camera in front of the vehicle detects the license plate of the front of the vehicle and starts shooting. The shooting content includes the license plate of the front, body and rear of the vehicle. When the license plate of the rear of the vehicle is captured, the shooting stops; S2, performing image recognition based on the license plate images of the front and rear license plates to obtain the license plate number, performing image stitching based on the image of the vehicle body to obtain a complete vehicle side panoramic image, and storing the license plate number in association with the vehicle side panoramic image; S3. Based on the vehicle side panoramic image formed by image acquisition, a data set for deep learning algorithm training is established, and a feature extraction and feature matching model is formed after training, and a feature matching algorithm is formed; S4, deploying the trained feature extraction and feature matching model to the backend server, and forming a post-processing logic program of the feature matching result based on the feature matching algorithm, wherein the post-processing logic program includes normalization processing and similarity score calculation, and performing normalization processing and calculating the similarity score on two vehicle side panoramic images obtained when the same vehicle enters and leaves the factory area; S5. Based on the post-processing logic program, a similarity score is obtained, and the similarity score is compared with a set threshold. If the similarity score is greater than the similarity threshold, the vehicles are judged to be the same vehicle; if the similarity score is less than the similarity threshold, the vehicles are judged to be different vehicles, and an alarm is issued.

2. The anti-cheating method for transport vehicles according to claim 1, characterized in that: In step S1, the camera is a three-eye camera, which respectively captures the front, rear and body of the vehicle, or the number of the cameras is three, which are responsible for capturing the front, rear and body of the vehicle respectively.

3. The anti-cheating method for transport vehicles according to claim 2, characterized in that: A marking spraying device is also provided outside the weighing platform, and the marking spraying device is on the same side as the camera. The marking spraying device sprays a detection mark on the side of the vehicle compartment, and the detection mark is photographed by the camera when the vehicle enters the weighing platform.

4. The anti-cheating method for transport vehicles according to claim 3, characterized in that: In step S1, the vehicle stops in front of the weighing platform, the marking spraying device approaches the side of the car and sprays paint at any position on the side of the car according to the image determined by the back-end server to form a detection mark.

5. The anti-cheating method for transport vehicles according to claim 3, characterized in that: The detection mark is a plurality of parallel transverse stripes arranged from top to bottom. On the left side of the detection mark, the ends of the plurality of transverse stripes are flush. On the right side of the detection mark, there is a length difference between the ends of adjacent transverse stripes. The plurality of transverse stripes form a plurality of length differences, and the plurality of length differences are different.

6. The anti-cheating method for transport vehicles according to claim 5, characterized in that: In step S4, all length differences in the same detection mark are processed to eliminate the influence of the shooting angle.

7. The anti-cheating method for transport vehicles according to claim 6, characterized in that: The method of processing the data of all the length differences in the same detection mark is: calculating the ratio of all the length differences, or measuring and calculating the length differences after processing the panoramic image of the vehicle side.

8. The anti-cheating method for transport vehicles according to claim 3, characterized in that: The sprayed pigment is a self-degradable pigment.

9. A transportation vehicle anti-cheating system, characterized in that: The anti-cheating method for transport vehicles is applicable to any one of claims 1-8, comprising a control module, a camera and a marking spraying device, the marking spraying device comprising a manipulator, a paint tank, a pump and a nozzle, the camera is arranged outside the vehicle entry end of the weighing platform, along the vehicle entry direction, the marking spraying device is arranged in front of the camera, the vehicle first passes the camera and then the marking spraying device and then enters the weighing platform, the camera, the manipulator and the pump are all connected to the control module, the pump is respectively connected to the paint tank and the nozzle through pipelines, and the pump is arranged on the manipulator.

Images