A segmentation algorithm based on an iterative computation of the mean shift filtering

Image segmentation is accepted to be one of the most important problems in image analysis. The good performance of any recognition system strongly depends on the results provided by the segmentation module. According to many researchers, segmentation finishes when the goal of observer is satisfied. Experience has shown that the most effective methods continue to be the iterative algorithms. However, a problem with these algorithms is the stopping criterion. In this work, we present a strategy for image segmentation through a new algorithm based on recursively applying the mean shift filtering, where entropy is used as a stopping criterion. The main feature of the proposed algorithm is to carry out segmentation in an only step. In other words, with the new algorithm is not necessary to carry out additionally the segmentation step, where in many occasions (mainly in complex applications), it can be computationally expensive. The effectiveness of the proposed algorithm is shown through several experimental results. The obtained results proved that the proposed segmentation algorithm is a straightforward extension of the filtering process. In this paper a comparison between our algorithm and so called EDISON System was carried out.