Segmentation of colored digital images using region based approach

Abstract EN

Image segmentation is a low-level operation intended to partition the image into homogeneous disjoint regions.

Many algorithms has been developed by the scientific community, for the grayscale image segmentation.

However, the colored image segmentation problem, which is quite more important, has received much less attention in the literature.

The main contribution of this paper is to improve the seeded region growing algorithm so that the algorithm can perform the segmentation operation automatically without the need for human supervision.

Our algorithm, also, analyzes several other features of images, such as the shadow, light reflection and intensity.