A Multi-landmarks Localization Method Of Mobile Robot Based On Omni-directional Vision

Localization is an important issue in the study of the mobile robot and its utilization. It is the prerequisite to realize robotic intelligence. a localization method based on multi-landmarks with the vision system of the robot is implemented here. The method makes use of the omni-directional camera which is fixed on the top of the robot to capture the surrounding images and uses the image processing technology to recognize landmarks, then determine the orientation and distance of the camera relative to these landmarks. So the location of robot can be gotten.The paper expounds localization of the mobile robot and correlative technologies, analyses several kinds of panoramic image transformation, graying, smoothing, sharpening, contrast stretching, edge-detection and so on, through analyzing the character of the panoramic image to gain the threshold of these transformations, and several kinds of edge-detection algorithm are compared when used to the panoramic image, finally the Canny algorithm is chose.Because Indoor mobile robot does curvilinear motion of changing steering, a method of line-landmarks extracting is implemented here, firstly extracts the possible line roughly, then does the rupture processing with the length constraint to gain the final effective line-landmarks.There arc some shortcomings for common camera, such as limited visual field. So the omni-directional camera is introduced. Its model, design and restriction are presented in details, The natural characters such as door, water pipe and institutional board are gained by feature extraction in omni-directional picture, these characters are the line-land marks, although some effective line-landmarks may not be extracted because of the light field, air flow, surface smooth of reflector, but robot localization using trigonometry is experimentalized with the residual effective line-landmarks. The result of experiment shows that this localization method is viable.In the final of the paper, the research and results are summarized. The future of this research field is described.