A SLIC based camouflage region selection method for mobile patrol robots

Abstract

Patrol robots are often in dangerous situations and may be damaged by criminals. To deal with this problem, a method to protect the robot by camouflage is proposed. It divides scene regions based on color, texture, and spatial features to search for suitable hiding regions. SLIC is employed to pre-divide scene images, in which three different image features are used to describe superpixels. Furthermore, superpixels with similar characteristics are classified as a same category. It can get a region by merging a kind of superpixels. Several scenes are applied to test the proposed method. The experimental result shows that the proposed method can realize camouflage region selection for patrol robots to hide.