Image Segmentation by using Mean-Shift with Dynamic Region Merging - A Survey
Shahab Ahmed, Zeeshan Khan, Anurag Kumar Jain · International journal of advanced research in computer science and electronics engineering · 2012
Image segmentation is to classify or cluster an image into several parts (regions) according to the feature of image. This paper presents a survey of different image segmentation techniques. Efficient and effective image segmentation is an important task in computer vision and object recognition. Since fully automatic image segmentation is usually very hard for natural images, interactive schemes with a few simple user inputs are good solutions. In this paper, several segmentation techniques using different segmentation algorithms like Watershed, graph based, mean-shift etc with dynamic region merging approach are taken from the literature are reviewed. In this paper, the main aim is to understand that which algorithm works best with dynamic region merging for image segmentation.