A NOVEL FUZZY BASED BACKGROUND SUBTRACTION FOR DYNAMIC SCENES

Detection of moving objects in video streams is the initial relevant step of knowledge in several computer vision applications. Background Subtraction is widely used as the basis for moving object extraction and tracking in video sequences. For this, most of previous strategies rely on the idea that the background is static over short period of time. On the other hand, video sequences with dynamic background such as waving leaves, rippling water are not easy to process and it leads to poor extraction of foreground objects. In this paper, a completely unique thin Background Subtraction algorithmic program for scenes with dynamic background is planned to solve this drawback, which uses fuzzy color histogram in which Fuzzy C Means Clustering (FCM) module is based on correlation method and this can also be applied to medical image applications. Experimental results give substantiating proof that the proposed technique is effective for Background Subtraction to track the moving object in Dynamic texture scenes.