Research On The Key Technology Of Road Edge Recognition Of Unmanned Truck Based On Machine Vision

With the rapid development of computer technology,intelligent mining machinery will be one of the development directions of intelligent mine.In this paper,aiming at the problems of fuzzy road edge and low recognition rate under the complex environment of mining area,the image processing technology in the mining area unstructured road edge recognition and monitoring is studied in depth.The main work of this paper is as follows:(1)Based on the feature analysis of the road image in the HSV color space in the opencast mining area,the road image is preprocessed in the HSV color feature space to reduce the influence of light and the redundancy of system calculation.On this basis,an image enhancement algorithm based on Retinex theory is proposed.(2)According to the characteristics of unstructured road edge shape in mining area,a fitting method based on road left and right edge segmentation is proposed.Three fitting models of straight line,curve and straight parabola are constructed.Finally,the improved Kalman filter is used to track and detect the road image.(3)Based on the MATLAB platform,the paper analyzes the implementation effect of the road edge detection,edge fitting and edge tracking algorithm proposed in the paper to process different types of road images in opencast mining area,and uses three evaluation indicators to objectively evaluate the application of the algorithm.This paper analyzes the applicability and robustness of the algorithm for the road edge of open-pit mining area with high unstructured degree.