An algorithm of moving objects detection for indoor environment based on laser and vision

Abstract

In this paper, a new algorithm of moving objects detection is proposed for indoor environment, which can effectively overcome light changing and can better detect and extract moving objects in indoor environment. The algorithm first transforms images in the form of RGB into the form of HSI for better image processing, and then detects and extracts moving objects based on image and distance information. One core of the algorithm is effectively combining information of image and distance during the process of motion detection, which can better overcome light changing in indoor environment; the other is reasonable using of the background subtraction and the frame difference during the process of moving objects extraction, so that the moving objects can be extracted more accurately and completely. Experimental results have validated the feasibility and reliability of the algorithm.