Home Service Robot Intelligent Space Key Technology Research And Design,

The increasing development of robot techniques makes it possible to devise and develop service robots for civil applications. However, general indoor environment can be variable and disturbed frequently, which bring about much difficulty for research and development of home service robots. Intelligent space is a promising technology, which is expected to help home service robot wok in complex environment more easily and effectively. Intelligent space is builded by setting many kinds of sensors and manipulators in the indoor environment. It gets the environment information, and then instructs the robot to serve the users.We focus on the following key technologies for service robots to work in the intelligent space and finally present an outlook on the development of the intelligent space techniques.The artificial landmark technology is summarized and a new kind of artificial landmark is proposed that is planar self-similar landmark based on OR code technology, and its characteristic is introduced too. The method to detect and recognize this landmark is studied, and how to realize robot autonomous localization only by single landmark is researched. Experimental results show that the landmark can be detected and recognized accurately and the robot can locate fast and accurately using this kind of landmark in the complex dynamic environment. In addition, the idea that how to use RF tag in the intelligent space as a new kind of landmark is proposed.The panorama vision system using upper-projectors is researched. The image is pretreaded and segmented by edge connection technology, then is recognized using gaussian mixture model based on color information, so that the environment model is constructed. Experimental results show that the objects in the indoor environment can be recognized with high accuracy, and the environment model can be constructed correctly. A new project of straight lines extraction is proposed, and the camera intrinsic parameters is calculated by using self-calibration method based on three vanish points, and then the objects in the environment is located using camera extrinsic and intrinsic parameters.Image mosaic technology is used in this paper in order to model and locate the panorama image. The feature points of every images are extracted firstly by the SIFT algorithm, and then matched by the Euclidean Distance algorithm and the RANSAC algorithm. The mosaic image is acquired based on the matched feature points.Finally, the limitations of this paper are pointed out, and some ideas about the furture development of the intelligent space are proposed.