Distinguishing The Dated Marker Lines Of Highway Based On Image Processing

The highway mileage increased rapidly and more and more labours are employed to maintain the highways due to the lackage of the marking vehicle with higher degree of automation. It is necessary to study and develop the marking vehicle with high efficiency and intellectualization for improving the mechanization level of road maintenance and reducing the labor intensity of workers in china. In the marking, it is difficult to automaticly distinguish the dated marker lines and cover a new one on them. Automaticly distinguishing the dated marker lines is the key technical problem to urgently solve. Therefore, this study focuses on automaticly distinguishing of the dated marker lines using machine vision and image processing. The main research work is as following:1. The identification of the dated marker line of road was discussed. A machine vision system was developed to collect the marker lines images. Gray-scale transformation (GST) and histogram analysis approaches were discussed;2. Smoothing and median filter on the images of the marker lines were discussed. A novel dynamic filtering algorithm was proposed by analysis the characteristics of the image of dated marker lines. The results showed that the proposed method performed better than the traditional smoothing and median filtering when the dynamic algorithm was used to remove noise of road marking image;3. After road marking images were binarized, morphological operations and edge withdrawing were conducted on the binary images. The least squares was used to simulatebthe centerline of marker lines. Finally, the center line was used to guide and control sprayer. The results showed that the accuracy of the simulated centerline of the road marking could satisfy the requirements of identifying the dated road marking and covering a new road marking on it;4. The Study on the calibration method of the camera.was carried out. The camera of machine vision system was calibrated with the polynomial algrithem. The results showed that the method is simple, practical, high efficiency, and the obtained results were satisfactory. and could meet the practical requirements for the road marking.