| To solve the problems of low efficiecy and high labour intensity,with the cold milling machine when it is working and moving, which is absolutely manipulated by operator,and to highten the level of automation and intelligentization of the road milling machine,the technology of machine vision applied to work navigation of road milling machine is proposed in this paper. The technology and methodology in this reseach can be adopted to other go-working machinery.The systems frame of vision navigation milling machine and vision navigation algorithm have been designed in this paper.Also,the road image processing have been researched. Thereinto,the vision navigation algorithm is emphasised on solving the problems that three-dimensional information is reflected from the two-dimensional image information and how to get navigation parameters. The purpose of discussing image processing is to find the adaptive method to achieve road image processing,so that the navigation parameters could be caculated. In the term of vision navigation,the map from two-dimensional image information to three-dimensional information can be achieved by increasing constraint condition. In the term of image processing,to find the adaptive method to achieve road image processing many procedded results using all kinds of image processing have been contrasted. At the same time,the image binaryzation is done before the image filtering,so that the calculation is reduced and the real-time performance is improved. In the end,for some characteristics of the vision navigation milling machine such as complicated model,uncertain work envirenment, nolinear and hysteresis quality, fuzzy control has been used to make the vision navigation milling machine to work following the navigation datum line. The fuzzy controller of vision navigation has been designed.The controller performance has been simulated using the MATLAB software. |
