Semi-supervised Image Segmentation Based on K- means Algorithm and Random Walk

Abstract

Semi-supervised image segmentation is a process of classifying unlabeled pixels using known labeling information. In order to realize image segmentation, solve the problem of setting a large number of seed points in the random walk algorithm, and solve the local optimization problem in the K- means algorithm, this paper proposes a semi-supervised image segmentation algorithm based on the K-means algorithm and random walk. Firstly, the K-means algorithm is used for clustering to determine the clustering center, then, the transfer probability from each unlabeled pixel to the seed point is calculated based on the random walk algorithm, and the image segmentation is completed according to the transfer probability. It can be seen from the experimental results that the segmentation accuracy is greatly improved and the effectiveness of this paper is verified.