Font Size: a A A

Research On Lane Detection Technology Based On Machine Vision

Posted on:2020-02-21Degree:MasterType:Thesis
Country:ChinaCandidate:K GuoFull Text:PDF
GTID:2392330572482470Subject:Instrumentation engineering
Abstract/Summary:PDF Full Text Request
The military's shooting training method is mainly for target training,and the existing target machine has the problem of inflexible movement and single movement direction.Therefore,a smart and sensitive smart target vehicle is designed,which can greatly improve the shooting level of the soldiers.At present,the positioning technology used by the smart target vehicle may cause inaccurate positioning when moving along a specific route indoors,causing a problem of oscillating movement.In this paper,based on the above problems,the research on the detection technology of intelligent target vehicle lane line based on machine vision is carried out.The main research work of this paper is as follows:1.In view of the problem that the dynamic acquisition lane image is overexposed.the method of adjusting the camera shutter and gain based on the histogram gray distribution control is adopted to improve the image contrast.For random noise and uneven illumination in the image,median filtering and bottom cap transformation are used to improve the local features of the image.In order to extract the edge of the lane line quickly and accurately,the image is scaled by bilinear interpolation method to speed up the image detection speed;then the maximum inter-class variance method is used for threshold processing;finally,the Laplace operator is used to extract the lane line edge.2.For the lane line fitting detection problem,the lane center line is obtained by the mean method,and then the least square method is used to fit the continuous lane reference line.In order to improve the detection efficiency of the lane line,based on the idea of Kalman dynamic tracking,the dynamic prediction of the area where the lane line is located is proposed,and then the lane line detection is performed in this area3.The vehicle control information is obtained by the fitted lane reference line,and the control information is transmitted from the PC to the single chip microcomputer.and the PWM pulse width is controlled by the incremental PID algorithm to increase the stability of the lane line detection.The function modules of the detection system are written in C and C++ language,and then the modules are connected together by Qt interface for display.The results show that the edge detection algorithm used in this paper can quickly and accurately extract the edge of the lane line,has good anti-interference performance,and meets the requirements of edge detection;the correct rate of dynamic area detection is increased by 4 percentage points to 96.4%,and the average time is Reduced to 45ms,meeting the requirements of real-time detection.Through the operation of the detection system,it is concluded that the lane detection system can successfully control the trolley to travel along the indoor lane line.
Keywords/Search Tags:Machine Vision, Lane Line Detection, Kalman Prediction
PDF Full Text Request
Related items