Edge, Shade and Mixed Range Detection by Fuzzy Gaussian Filter for an Autonomous Robot

Extraction of edge, shade and mixed ranges are the first stage in the process of image understanding. This paper highlights the difficulties for a mobile robot to find out these features from a video stream of images transmitted from the robot. As the robot moves through the uneven lighting condition in its surrounding, the image quality differs from point to point. The paper proposes a recursive formulation of fuzzy Gaussian filter, by varying its parameters, i.e., window size and σ value to get the dominant features of the online video images. The method has been implemented on a client machine for the video images transmitted by the mobile robot Pioneer DX-2. The result shows that the method is suitable for the online images taken in varying lighting conditions.