Object segmentation using maximum neural networks for the gesture recognition system

In this paper, we present a new clustering method for segmentations of moving target and non-target objects. We assume that the moving target object has the following conditions: (1) object motion data continuity inter-frame, and (2) object motion data continuity intra-frame. In our model, clusters tend to form as 2lling these two conditions. The experimental results showed the e3ectiveness of the proposed algorithm and the performance of this model in terms of the quality of the recognition results. Our algorithm is able to clean the input noise by removing non-target objects before the recognition process. c © 2002 Elsevier Science B.V. All rights reserved.