A Multiscale Block-based Color Segmentation Technique

X. Lu and B. Yang (PRC)


Contours in textile color images, edgedetection, multiscale segmentation, multiresolutionsequential edge linking


This paper studies the problem of combining a multiscale block-based segmentation and the multiresolution sequential edge linking for finding contours in textile color images. First to eliminating the disturbance of "texture noise", the dominant colors in a textile color image are extracted according to the human observation, The multiscale approach is used to solve the conflict between boundaries localization and high resolution segmentation, and the segmentation result is recursively refined at each scale based on a contextual model. Experimental results show that our algorithm can achieve better segmentation results when tested on color textile images, and also produce high quality edge images when compared with the mean-shift algorithm [1] and the spatial gradient operator proposed in [2].

Important Links:

Go Back