| Because of the limitation of its own function,ground mobile robots have limited perception of the global environment.Although the drones can observe the global range of the ground environment at a specific altitude,their load and endurance capacity are relatively lacking.Therefore,the space robots cooperation system has complementary advantages.This paper builds an air-ground robot collaboration platform with NAO robots and drones.The main research contents of this paper are as follows:In the detection of drones to the ground,due to the detection bias caused by the ground shadows and other influences when drones perceive the ground environment at high altitude,a method of detecting the target with color features and contour recognition is proposed.In the path processing,we use the combination of image processing methods to eliminate the irrelevant features and extract the complete path.The method of thinning and comer feature detection is used to achieve extracting of the corner points,and complete the ground search needs.In the problem of path planning,because the generated ground path is a two-dimensional array of mazes,the breadth-first algorithm is selected as the path planning to perform navigation by comparing the more commonly used two path planning methods,depth-first and breadth-first.In the communication combination of open space robots,wireless broadband is selected as a suitable communication method according to the actual conditions of NAO robots and drones,and corner point data is encoded in combination with the cormer feature to realize the communication of open space robots.In the ground robot performing ground missions,the NAO robot uses the Hough transform for heading correction and side-track correction.In target recognition and grasping,after calculating the position of the target in the image,we use the kinematics principle to complete the grasp of the stick.Finally,the program was written in Python and verified experimentally. |
PDF Full Text Request