The obstacles detection for outdoor robot based on computer vision in deep learning

In this thesis, we propose a guided robot with deep learning techniques and machine vision in order to predict distance and avoid obstacles. The training data images are captured by the stereo camera, and output results are disparity of the single image. Then, the distances are converted by using the triangulation method of computer vision. The back propagation neural network retrains to obtain the actual distance of each pixel in the image. Therefore, the robot with the monocular camera could know the distance between obstacles and itself. Semantic segmentation is utilized to a to distinguish road and obstacles in the image. Fuzzy theory for calculating the area of the road which be cut is designed to avoid walking into the intersection. The obstacle depth and the walkable area are taken as an information for avoiding obstacles control. Because of the robot is proposed to walk on the right side along the road, the edge of road is necessary to navigate the robot. The Hough method is used to find the straight line, and then choose the line we need. The path planning is also achieved in through of navigation and obstacle avoidance control. As a result, the robot can arrive the destination safely and precisely.