The visual-servo control technology of mobile robot which uses the visual sensors to access environmental information and controls the mobile robot to reach specified posture or track the target object is one of the hotspot in the field of mobile robotics research. Apart from using the classic visual-servo control technology which takes industrial robot as the research object, the nonholonomic constraints introduced by wheels must be taken into consideration which impacts on the control of mobile robot. The paper investigates on visual-servo control technology of wheeled mobile robot in the indoor environment, and verifies the effectiveness and robustness of the proposed algorithms through related experiments. The main contents and innovations of the paper are as follows: Firstly, the visual-servo control technology based on artificial landmarks is studied. The colored adhesive tape and MR code which are pasted on the ground surface are used as the artificial landmarks. The recognition algorithm of these artificial landmarks is given and a posture calculation method of the mobile robot relative to the colored adhesive tape and MR code is proposed. With the visual feedback, the control law is designed to implement the trajectory tracking of the mobile robot. The visual-servo control technology based on artificial landmarks has been applied to the tractive AGV designed by our lab. Reliable running in the warehousing and logistics environment verifies the validity and reliability of the algorithm. Secondly, the visual-servo control technology based on natural landmarks is studied. SIFT feature is used as the natural landmark. Firstly, the paper presents an outer point elimination algorithm based on density clustering, which can improve the accuracy of SIFT features matching and reduce false matching. The scene identification and location of the mobile robot is realized by analyzing the matching information of the SIFT feature points between the current image and database images. Combining the SIFT feature and the depth image from Kinect camera, a docking method is presented in this paper and the effectiveness is verified through the relative experiment. Thirdly, the visual-servo control technology based on the depth image gotten from the depth camera is discussed. The Microsoft Kinect camera is used to obtain environmental depth information and an image preprocessing method is proposed firstly to deal with the depth image...
修改评论