Sensitive and accurate complex road condition lane deviation real-time early warning system

Abstract

With the practical application of developing an intelligent auxiliary driving system as the background, a top-speed lane deviation real-time early-warning system is designed and realized on a vehicle-mounted ARM platform, a top-speed lane line recognition algorithm for image W-r extension dynamic evaluation is provided, lane information is detected in real time, and the vehicle lane deviation early-warning system is obtained by considering the transverse speed of a motor vehicle in a lane. Two early warning strategies are combined, in the first mode, when the transverse speed is low, the convenient relative distance between a motor vehicle and a lane is adopted for early warning judgment, and under the condition that the transverse speed is high, lane departure judgment is conducted according to the deviation rate of a left lane line and a right lane line in an image; according to the method, perspective transformation of the image is not needed, external parameters and internal parameters of the monocular camera do not need to be measured, the requirement for the monocular camera is low, and great convenience is brought to a user. And meanwhile, the algorithm is low in time complexity, high in speed and high in practicability and applicability, and has huge value and wide market in real-time early warning of lane departure under complex road conditions.

Description

technical field [0001] The present application relates to a real-time early warning system for lane deviation in complex road conditions, in particular to a sensitive and accurate real-time early warning system for lane deviation in complex road conditions, which belongs to the technical field of lane deviation early warning for intelligent assisted driving. Background technique [0002] The rapid development of the automobile industry has also resulted in a gradual increase in traffic accidents. Various complex terrains have led to complicated road conditions. At the same time, the quality of drivers is uneven. Traffic accidents are prone to occur, and traffic safety is facing a serious test. [0003] In order to avoid car accidents and improve driving safety and comfort, intelligent assisted driving has become popular, and vehicle safety assisted driving systems have become a research and development hotspot. Among them, the lane departure warning system is an important pa...

AI Technical Summary

Problems solved by technology

[0008] First, through the classification and analysis of the existing lane detection algorithms, it is concluded that there are some or other deficiencies when they are directly applied to the vehicle lane departure warning system; although the lane detection algorithm based on the basic features of the image has low calculation , but its robustness is poor, and many complex road conditions cannot be detected or false detection occurs; although the lane line detection based on geometric information is better, it involves calculation modules such as perspective transformation, which requires a large amount of calculation, and Many parameters need to be adjusted, which is difficult to realize in real time; the lane line detection model based on color information is better for simple road conditions, and the speed is fast, but it is difficult to use under various complex road conditions; the lane line detection model based on deep learning Line detection is a relatively new field. Although deep learning has demonstrated powerful capabilities in the fields of big data processing and face recognition, its platform requirements are high, and there are still many problems in its real-time performance and effect in lane line detection. The existing technology lacks a real-time warning algorithm and system for lane deviation with fast speed, good effect, practicability and applicability;

[0009] Second, the prior art lane departure warning hardware platform is not economical and practical, and the hardware cost is high, and the stability and real-time performance of the vehicle-mounted system are poor, and there are often obvious lags and even failures, which cannot be effectively reduced. Traffic accidents occur, reducing casualties; at the same time, the cost of updating and using is high, low cost, narrow scope of application, large system size, inconvenient installation, weak power consumption and poor economy when used on motor vehicles. Although there are various detection algorithms, the amount of calculation is relatively large, and the detection speed is too slow to be used in the vehicle system. Real-time lane line detection and early warning cannot be realized on the vehicle equipment. Therefore, the development of a lane departure warning technology based on the ARM platform has Great significance and great practical value;

[0010] Third, the prior art lacks a fast lane line recognition algorithm that can meet real-time requirements, lacks a fast lane feature optimization modeling method, and cannot detect complex road condition lane information stably and quickly in combination with road features. The perspective transformation calculation module used by the algorithm has a large time complexity, which not only has little effect, but also greatly increases the amount of calculation when used in the vehicle system, slows down the detection speed, and cannot combine the edge point intensity and the parallel relationship feature selection of the lane edge. For lane lines with high reliability, it is impossible to estimate the parameters of angle, offset, edge average intensity, and custom curvature, and it is impossible to use lane color information to distinguish yellow and white lane lines. There is no method to distinguish virtual and real lane lines by edge point spacing; There are technical methods that can only detect the straight line of the lane and cannot reflect the characteristics of the curvature of the lane. It requires the conversion of each coordinate system. In engineering applications, it is necessary to obtain the internal and external parameters of the camera, the installation location, and the vehicle model. The extraction degree of the lane edge is too low , it is difficult to realize real-time detection of lane lines on the ARM platform with weak computing power;

[0011] Fourth, after accurately determining the position information of the lane line, it is necessary to obtain the relative position information of the motor vehicle, and calculate the various possibilities that the driving motor vehicle will deviate from the lane through the deviation strategy; the existing technology is based on the lane model and real road surface coordinate information. The adaptability of the early warning method is weak. To establish the geometric imaging coordinate system of the motor vehicle system, camera, and road surface, it is necessary to know the model size of the motor vehicle, the type of road, the rotation angle and height of the camera and optical lens, the internal parameters and distortion parameters of the camera , every time a car, road model or camera is changed, the coordinate system must be re-established, which is very troublesome in engineering; at the same time, the camera needs to be calibrated after perspective transformation and other processing involving image pixel coordinate conversion, which is very large To a certain extent, the trouble of actual use is increased, and the calculation speed and operation simplicity are poor; the establishment of the road model to realize the deviation early warning is contrary to the original intention of the algorithm design, and the lateral speed of the motor vehicle in the lane is not considered, and the two early warning strategies are integrated. The method lacks the use of the convenient relative distance between the motor vehicle and the lane for early warning judgment when the lateral speed is not large; lacks the deviation rate of the left and right lane lines in the image for lane departure when the lateral speed is large Judgment, lane departure warning cannot achieve the desired effect

Images