A vehicle positioning system based on peripheral vehicle information and machine vision

Abstract

The present application belongs to the field of vehicle positioning and navigation, mainly consisting of information recognition module, image acquisition module, information sending module, information receiving module, addressing positioning module and data processing module. The system first determines the search radius according to the intersection Bluetooth on the target vehicle to search for the Bluetooth signal, and then compares and matches the color and width information of the vehicles collected by the intersection camera and the color and width information of the surrounding vehicles collected by the camera on the target vehicle to determine the position of the surrounding vehicles relative to the intersection. Taking the intersection Bluetooth as the center and the radius as the arc, combined with the position of the surrounding vehicles relative to the intersection, the position of the target vehicle relative to the intersection is preliminarily determined to complete the preliminary positioning. Then, the position of the target vehicle relative to the intersection is calculated through the camera in front of the intersection to complete the secondary positioning, further improving the positioning accuracy of the vehicle.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of vehicle positioning and navigation, and particularly relates to a vehicle positioning system based on surrounding vehicle information and machine vision. BACKGROUND

[0002] The automobile industry is a pillar industry of a country. With the improvement of the national economic level, the automobile ownership is also steadily increasing. Vehicle positioning technology is very important in vehicle safety and is an important part of traffic safety.

[0003] The existing vehicle positioning method and related device determine the position of the vehicle by means of a device with high positioning accuracy in the surrounding equipment to improve the positioning accuracy of the vehicle. However, the prerequisite is that the vehicle to be positioned has V2X communication capability and UWB electronic tag, which makes most non-networked vehicles unable to apply this technology, and has certain limitations. Or based on long and short term memory neural network and untraceable Kalman filter fusion of multiple sensors to improve the positioning accuracy of intelligent driving vehicles. However, the fusion of multiple sensors is relatively complex, and the price of laser radar is relatively high. On the other hand, GPS signals are easily affected by the external environment, and may not be able to normally search for GPS signals in high-rise buildings, under viaducts, in tunnels or in mountainous areas. Or a combined positioning system based on vehicle-road cloud cooperation, although it overcomes the defects of unstable and inaccurate positioning caused by the influence of complex special environment on GPS, but the system implementation is relatively complex.

[0004] The above vehicle positioning system is either only for the positioning of intelligent networked vehicles, or based on GPS technology positioning, or the system implementation is relatively complex. However, the intelligent networked vehicles running on the road only account for a very small proportion of road vehicles, and most of them are ordinary non-networked vehicles. GPS signals are also affected by complex environments, resulting in inaccurate and unreliable positioning. In view of the defects of GPS positioning and the high implementation cost of positioning method based on fusion of laser radar and other sensors, the present application proposes a vehicle positioning system based on surrounding vehicle information and machine vision. SUMMARY

[0005] In order to realize the positioning of the target vehicle in the GPS-free environment, the vehicle positioning system based on surrounding vehicle information and machine vision comprises an information recognition module, an image acquisition module, an information sending module, an information receiving module, an addressing positioning module and a data processing module.

[0006] The functions, effects and implementation processes of each module are briefly described as follows.

[0007] The information recognition module is used to collect the physical information of the surrounding vehicles, including color and body width. The color and width information of the front vehicles a1 and a2 of the target vehicle b are recognized by the two monocular cameras 1 and 2 arranged behind the front windshield of the target vehicle b, so as to realize the information recognition of the surrounding vehicles by the target vehicle.

[0008] The image collection module is used to collect the color and width information of the surrounding vehicles in front of the target vehicle, and collect the imaging projection of a point on the target vehicle. The image of the target vehicle and the surrounding vehicles is collected by the monocular camera 3 arranged at the intersection with a distance d from the target vehicle b, a height h from the ground, a pitch angle θ and a focal length f.

[0009] The information sending module is used to send the color and width information of the front vehicles a1 and a2 recognized by the information recognition module to the information receiving module. The information sending module is realized by the element Bluetooth 1 with an id1 address arranged on the target vehicle b.

[0010] The information receiving module is used to receive the color and width information of the surrounding vehicles sent by the target vehicle b, and identify the id address of the Bluetooth on the target vehicle b. The pairing of the information sending device Bluetooth 1 is completed by the element Bluetooth 2 with an id2 address arranged at the intersection, and the information transmitted by the Bluetooth 1 is received.

[0011] The addressing positioning module is used to search and match the Bluetooth 1 on the target vehicle by the Bluetooth 2 at the intersection, to determine the position radius of the target vehicle relative to the intersection, and to determine the relative position of the surrounding vehicles of the target vehicle according to the color and width information. The signal of the id1 of the Bluetooth 1 on the target vehicle b is determined by the Bluetooth 2 with the id2 address at the intersection, and the distance r of the Bluetooth 1 is determined by the Bluetooth 2, to form a circular arc search area with the Bluetooth 2 at the intersection as the center and the search radius r. The color and width information of the surrounding vehicles collected by the camera 3 at the intersection and the color and width information of the surrounding vehicles collected by the cameras 1 and 2 on the target vehicle are compared and matched, to determine the position of the surrounding vehicles relative to the intersection. The position of the target vehicle relative to the intersection is determined within a certain error range by the circular arc with the Bluetooth 2 at the intersection as the center and the radius r, and the position of the surrounding vehicles relative to the intersection, to realize the rough positioning of the target vehicle.

[0012] The data processing module is used to calculate the distance between the target vehicle and the intersection camera 3. By projecting the image of the target vehicle on the monocular camera 3 at the intersection, the horizontal distance d between the intersection monocular camera 3 optical axis passing through the target vehicle b and the intersection of the ground is calculated to determine the position of the target vehicle.

[0013] The horizontal distance d between the intersection camera 3 and the target vehicle b is calculated by the imaging plane coordinate system origin (x0, y0), the coordinates of the target vehicle and the ground on the imaging plane (x, y), the height of the camera optical axis center from the ground h, the camera focal length f and the camera pitch angle θ. The calculation is as follows:

[0014]

[0015] By addressing the position to determine the position of the target vehicle, the preliminary positioning is completed, and then the horizontal distance d between the intersection camera 3 and the target vehicle b is calculated according to the intersection camera 3 and the target vehicle b. The intersection camera completes the positioning of the target vehicle again, and the positioning accuracy is improved. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 The flowchart for vehicle positioning is shown in the figure.

[0017] Figure 2 The system diagram for vehicle positioning is shown in the figure.

[0018] Figure 3 The system diagram for the intersection camera to calculate the horizontal distance between the target vehicle is shown in the figure. DETAILED DESCRIPTION

[0019] A vehicle positioning system based on surrounding vehicle information and machine vision includes an information recognition module, an image acquisition module, an information sending module, an information receiving module, an addressing positioning module, a data processing module, as shown in Figure 1 and Figure 2 as shown,

[0020] The information recognition module is used to collect the physical information of the surrounding vehicles around the target vehicle, including color and body width. By arranging two monocular cameras 1, 2 behind the front windshield panel of the target vehicle b, the color and width information of the front vehicles a1, a2 around the target vehicle b are recognized, so as to realize the information recognition of the surrounding vehicles by the target vehicle.

[0021] The image acquisition module is used for acquiring the color and width information of the surrounding vehicles except the target vehicle in front, and collecting the imaging projection of a point on the target vehicle.

[0022] The information sending module is used for sending the color and width information of the front vehicles a1 and a2 identified by the information identification module to the information receiving module.

[0023] The information receiving module is used for receiving the color and width information of the surrounding vehicles sent by the target vehicle b, and identifying the id address of the Bluetooth on the target vehicle b.

[0024] The addressing positioning module is used for searching and matching the Bluetooth 1 on the target vehicle b by the Bluetooth 2 at the intersection, judging the position radius of the target vehicle relative to the intersection, and judging the relative position of the surrounding vehicles of the target vehicle according to the color and width information by the camera 3 at the intersection.

[0025] The data processing module is used for processing and calculating the distance between the target vehicle and the intersection camera 3.

[0026] The vehicle positioning system based on the surrounding vehicle information and machine vision has the specific working process as follows:

[0027] Step1, the information recognition module recognizes the physical information of the vehicle in front of the target vehicle. The information recognition module is a two monocular camera 1, single 2 arranged on the target vehicle b, which recognizes and records the color and width information of the front vehicle a1, a2.

[0028] Step2, the information sending module sends the color and width information of the vehicle in front of the target vehicle recorded by the information recognition module to the information receiving module.

[0029] The information sending module is a Bluetooth 1 arranged on the target vehicle, with an ip1 address.

[0030] The information receiving module is a Bluetooth 2 arranged at the intersection, and the Bluetooth 2 also serves as an addressing positioning module, with an ip2 address.

[0031] Step3, the Bluetooth 2 with ip2 address as the addressing positioning module, addresses the Bluetooth 1 with ip1 address on the target vehicle b to determine the target vehicle signal. The Bluetooth 2 judges the distance r from the Bluetooth 1, forming a circular arc search area with the intersection Bluetooth 2 as the center and the search radius r.

[0032] Step4, by comparing and matching the color and width information of the surrounding vehicles sent by the target vehicle b with the color and width information of the vehicles in front collected by the intersection monocular camera 3, the position of the surrounding vehicles of the target vehicle relative to the intersection is determined.

[0033] Step5, with the intersection Bluetooth 2 as the center and the radius r of the circular arc, combined with the position of the surrounding vehicles relative to the intersection, the position of the target vehicle relative to the intersection can be determined within a certain error range, realizing the preliminary positioning of the target vehicle.

[0034] The image acquisition module is a monocular camera 3 arranged at the intersection, which functions to acquire the image of the target vehicle in front and send the image to the data processing module.

[0035] The data processing module functions to process and calculate the distance between the target vehicle and the intersection camera 3. By imaging projection of the target vehicle on the intersection monocular camera 3, the horizontal distance d from the intersection monocular camera 3 optical axis passing through a point of the target vehicle b to the camera plumb surface is calculated to determine the position of the target vehicle.

[0036] Step6, the image acquisition module sends the collected image of the target vehicle b to the data processing module, and the data processing module obtains the horizontal distance between the intersection camera and the target vehicle through calculation and processing, completing the positioning of the target vehicle by the intersection camera again.

[0037] Step 7, in combination with the position of the target vehicle relative to the intersection obtained by the preliminary positioning and the horizontal distance d between the intersection camera and the target vehicle, the target vehicle can be positioned accurately within a certain error tolerance. Through preliminary positioning and secondary positioning, the positioning accuracy is improved, and the positioning of the target vehicle is realized.

[0038] The calculation method of the horizontal distance between the intersection camera 3 and the target vehicle b is as follows:

[0039] As shown in Figure 3 , set the imaging plane coordinate system origin as P0(x0, y0), the coordinates of the intersection of the target vehicle and the ground on the imaging plane as P(x, y), the focal length of the camera 3 as f, the camera pitch angle as θ, and the height of the camera optical axis center from the ground as h. Through calculation, the horizontal distance between the target vehicle b and the intersection camera 3 can be obtained

[0040] It should be noted that the positioning of the address positioning and the calculation of the distance between the target vehicle and the intersection camera does not have a certain order. In addition to the steps described above, the horizontal distance between the target vehicle and the intersection camera can be calculated first to complete the preliminary positioning, and then the secondary positioning can be completed by matching the search signal and comparing the color and width information of the target vehicle and the surrounding vehicles.

Claims

1. A vehicle positioning system based on surrounding vehicle information and machine vision, characterized in that: The system consists of an information recognition module, an image acquisition module, an information transmission module, an information reception module, an addressing and positioning module, and a data processing module. The specific working steps are as follows: Step 1: The information recognition module identifies the physical information of the vehicle in front of the target vehicle; using two monocular cameras installed on the target vehicle, it identifies and records the color and width information of the vehicle in front. Step 2: The information sending module sends the color and width information of the vehicle in front of the target vehicle, which is recorded by the information recognition module, to the information receiving module; Step 2.1: The information transmission module is Bluetooth 1 installed on the target vehicle and has an IP1 address; Step 2.2: The information receiving module is Bluetooth 2, which is located at the intersection. Bluetooth also serves as an addressing and positioning module and has an IP2 address. Step 3: Bluetooth 2 with IP2 address is used as the addressing and positioning module. It addresses the information transmission module on the target vehicle, Bluetooth 1 with IP1 address, to determine the signal emitted by the target vehicle. Bluetooth 2 determines the distance r between itself and Bluetooth 1, forming an arc search boundary with Bluetooth 2 at the intersection as the center and a search radius of r. Step 4: By comparing and matching the color and width information of the surrounding vehicles sent by the target vehicle b with the color and width information of each vehicle in front collected by the monocular camera 3 at the intersection, the position of the surrounding vehicles of the target vehicle relative to the intersection is determined. Step 5: Using Bluetooth 2 at the intersection as the center and an arc with radius r as the center, combined with the positions of surrounding vehicles relative to the intersection, the position of the target vehicle relative to the intersection can be determined within a certain error range, thus achieving preliminary positioning of the target vehicle. Step 6: The image acquisition module sends the acquired image of the target vehicle b to the data processing module. The data processing module calculates and processes the image to obtain the horizontal distance d between the intersection camera and the target vehicle, thus completing the repositioning of the target vehicle by the intersection camera. Step 6.1: The image acquisition module is a monocular camera 3 placed at the intersection. The parameters are set as follows: the height of the center of the camera optical axis from the ground is h, the camera focal length is f, and the camera pitch angle is θ. Its function is to acquire images of the target vehicle in front and send the images to the data processing module. Step 6.2: Data processing module, whose function is to process and calculate the distance d between the target vehicle and the intersection camera 3; through the image projection of the target vehicle on the intersection monocular camera 3, calculate the horizontal distance d from the intersection monocular camera 3 optical axis through the intersection point b of the target vehicle and the ground to the vertical surface of the camera, and determine the position of the target vehicle; Step 6.3: The method for calculating the horizontal distance d between intersection camera 3 and target vehicle b is as follows: Let the origin of the imaging plane coordinate system be P0(x0, y0), the coordinates of the intersection point of the target vehicle and the ground on the imaging plane be P(x, y), the focal length of camera 3 be f, the camera pitch angle be θ, and the height of the center of the camera optical axis from the ground be h; the horizontal distance d between the target vehicle b and the intersection camera 3 can be calculated as . Step 7: By combining the position of the target vehicle relative to the intersection obtained from the initial positioning with the horizontal distance d between the intersection camera and the target vehicle, the target vehicle can be located relatively accurately within a certain error tolerance range. Through initial positioning and secondary positioning, the positioning accuracy is improved, and the target vehicle can be located.

Images