Study Of Object Location Algorithm For Automatic Charging Mobile Robot Base On Vision

With the development of the robotics, because of excellent autonomy and intelligence, autonomous mobile robot has become one of the hot topics in robot field and takes many kinds of tasks. When it comes to energy carrying capability of robotics, the length of working time is limited. The research of automatic charging technology has become the key to solve timing limit of mobile robotics, especially to interior environment. How to recognize the charging object is the difficulty of automatic charging technology. This paper has studied the technology of electric outlet location in interior environment.The research and development status of robot automatic charging is firstly reviewed, then introduce the autonomous mobile robot base on vision, include operational principle and system composition, also design the control strategy of robot automatic charging. A method base on area and figure is used to recognize two-phase sockets, and a kind of template matching method base on area is used to locate two-phase sockets. Considering the scaling, rotation and visual angle change, the thesis use SIFT arithmetic to search the SIFT feature points which match with electric outlet template from photograph for look for electric outlet object. Base on SIFT feature points have searched, This thesis also research the method to find the coordinate point has the same special location of electric outlet object in the left and right images which gained by stereo vision, this method compose by the process such as outliers elimination, spacial match points choice, coordinate compute. In the end, this paper experiment with binocular stereo vision system, gained the coordinate point of electric outlet object following this method. Then rebuild this point's three-dimensional coordinate in camera coordinate system. This method can gained the directional and depth information accurately.