Single image-based path planning for a spherical robot

With the working environments of the spherical robot become increasingly complex and unstructured, we must improve the autonomy of the robot. As a new mobile robot, the reliable navigation system and accurate sensory ability is very important. The paper presents a single image-based path planning algorithm for a spherical robot moving in an unknown environment. The robot explores the environment with the monocular visual sensor. Starting from the classic algorithms about the image preprocessing and edge detection, the close curves and edge points are extracted from the image. Then the points are fitted by least-squares in order to acquire the trajectory of the road boundary. On this basis, a safe local path can be planned out to navigate the robot. In the end, the practicability and validity of the method is verified by the result of the experiments.