Research On Road Obstacle Detection Based On Linear Structured Light

Line structured light measurement technology is widely used in industrial life,such as in the food industry,forestry industry,transportation industry,and electronics industry.This measurement method has the advantages of active and non-contact.The line structured light measurement technology is used in the line structured light vision measurement system.The system has a simple structure,is easy to acquire and process light bar images,and has good real-time performance.Compared with lidar,line structured light vision measurement system has the advantages of low cost and fast recognition in road obstacle detection and recognition,so line structured light measurement technology can be used to replace some functions of lidar to detect obstacles.The research content of this article is divided into the following parts:(1)Analyzed the working principle and mathematical model of the line structured light measurement system,and then constructed the detection system,selected the appropriate camera and line laser according to actual needs,and evaluated the performance,parameters and functions of the camera and line laser.(2)Studied the calibration of line structured light vision system.Use Zhang Zhengyou's checkerboard to calibrate the cameras in the system;compare and analyze the existing light plane calibration methods to find the shortcomings of the light plane calibration methods.A calibration method using a two-dimensional calibration plate is used at different heights to obtain the intersection of the light plane and the calibration plate.According to these intersecting light points,the light plane equation can be obtained,and then the light plane calibration is completed.At the same time,the surface information calculation formulas of different obstacles are derived.(3)Analyzed the existing algorithms for extracting the center of the light strips of the contrast line structure.According to the existing deficiencies of the existing algorithms,this paper combines a Gaussian fitting method and a gray center of gravity method to propose Gaussian weighted two-dimensional gray center of gravity method for extracting strips center.This method uses the gray-level neighborhood attribute method for rough extraction of the center of the light strips,which can remove the solitary noise and find the normal width value of each column of the light strips,so as to provide preparation for the subsequent accurate extraction of the center of the light strips.When accurately extracting the center of the light strips,based on the two-dimensional gray center of gravity method,the Gaussian weighting in the direction of the column can not only suppress the characteristics of the interference of the light strips trend,but also avoid the complex fitting process of the Gaussian fitting method.Experimental results show that the accuracy and real-time performance of the center of the light strips extracted by this method is better than the traditional gray-scale center of gravity method,direction template method and adaptive two-dimensional gray-scale center of gravity method in a complex environment,and has a strong stability and robustness.Through the assembly of the actual detection device and the measurement experiment under the complex background of the shade,the three-dimensional measurement map of the obstacle is obtained,and the maximum length,width,height or depth of the obstacle is obtained by using the information calculation formula.It proves the feasibility of this method,and has certain practicality.