Research Of Color Target Recognising And Tracking For Carrying Robot

As the same as human vision system, vision system is playing an important role in computer outside information collecting system. It is because computer vision system simulates the function of human eyes to collect outside information from image, and to recognize the configuration and feature of the goal objects. Color image can describe objects better than the gray-level image, so that the more information can be obtained from processing color image.The main contributions of the work include that a color target recognising and tracking and transit system is established, and then based on the theories and technologies about the segmentation of chromatic image, the motion control of mobile robot is studied.First of all, the color model is discussed to present the color recognising theory for robot. Some popular color space models, RGB and HSI, are introduced. The relation and difference between them are compared in both picture and data sheet. The image segmentation and clustering algorithm are discussed and used to detect objects in images. A method of using double threshold of H and S value to segment the object from the background is used in this paper. A fast scan line seed filling based on clustering algorithm for two-value image is proposed to get the objects' rectangle area. Then, kinematic model of the wheeled mobile robots are analyzed and the algorithm for motion control of wheeled mobile robot is presented finally, the target recognising and tracking experiment is carried out based on Pioneer 3-DX mobile robot.Experimental results show that the above algorithm is effective and fast operating.