195 Image Multi-Edge Detection and Texture Analysis for Line Guidance Visual Robot Navigation in Real World

Noise, uncertain and inhomogeneous environment have to be faced on very carefully when a line guidance navigating visual robot is put into a real world instead of ideal experimental conditions. These environmental factors sometimes are somewhat hardly to deal with. In this paper, we give a novel algorithm of multi-edge detection and image texture analysis to solve the problems resulting from noise and uneven illumination. To the visual based robot we discuss, the images from a charge coupling device camera in it are the only information to its controllers. This method make the robot catch the correct image information on line guidance while it is moving, for instance, from one starting room to some rooms of destination. This need could be found in some welfares hospitals and plants. The experimental results are satisfied. This approach is expected to give a practical way in all this kind of problems.