A Color Image Segmentation Method Based on Automatic Seeded Region Growing

摘要

We present a segmentation method based on SRG for color image. Seeds is automatically selected depending on calculating the pixel intensity difference of pixel in the Luv color space and relative Euclidean distances. Initial regions are developed by applying SRG to selected seeds and classified based on the region distance defined by the color spatial and adjacent information. The presented method can select the seeds automatically, which is unavailable in traditional ones. Moreover, it can avoid over- segmentation. The experimental results show that the proposed method is available to various color image segmentation.