| The application of mobile robot has gradually matured in military, medical, industrial and so on, with the development of science and technology as well as people living standard, robot replacing human to do labor with high risk, high-precision and mechanization has become a trend. Currently, there are a lot of ground smear such as collage and flyer, etc., and with its difficult cleared, recurring problems, which interferes with people’s normal life and the city’s appearance seriously. Therefore, this paper designs a robot for cleaning ground smear cleaning based on vision which can search target position and clean it autonomously on the ground.According to the actual demand of robot for ground smear cleaning, overall structure of system is designed. The paper takes robot structure analysis, the visual system modeling, image acquisition and processing and positioning obstacle avoidance algorithm as the research subject. Firstly, the overall structure of robot is introducted, and it presents every fuctional module and its hardware configurations. Moreover, the image on the ground is collected through monocular cameras on robot body for converting 3-d environment information to 2-d image information by monocular camera vision model, which is passed to the main control chip. Afterimage segmentation with dynamic threshold value method, it selects pixels calibration to extract image feature, and then actual position of ground targets is calculated. Finally, robot avoids obstacle autonomously using avoidance algorithm basis of the fuzzy control theory while encountering obstacle.Experiments show that the system can be effective on recognition through some steps such as image acquisition, processing and the calibration of characteristic pixel, and cleaning ground smear information according to sending commands to actuators by the control module with well cleaning. At the same time, by the obstacle avoidance algorithm designed in the paper, robot can avoid obstacles safely. |
PDF Full Text Request