Mean Shift Image Segmentation Based On NL-Means

Aim at the shortage of density central point choice in Mean shift image segmentation, this article uses a new way of searching for the density central point, in the meantime, in order to overcome the blemish of the space relation to the pixel of tradition feature space analytical ,during Non-local means algorithm ,using characteristic weight parameter in the distance formula ,thus make the clustering effect better .the experiment is shown that the proposed algorithm can be valid.