Chromosome Image Segmentation Based on OTSU and Region Growing Algorithm

Abstract

Chromosome image segmentation is of great significance for karyotype analysis. Traditional segmentation algorithms are widely used due to their fast speed and strong practicability. However, the traditional segmentation algorithm has the problem of incomplete segmentation of chromosome regions when processing chromosome images. This paper proposes a segmentation algorithm that combines Otsu and region growing algorithm. The region growing algorithm grows by a sliding window. The algorithm determines the growth criterion based on the improved Otsu threshold selection formula. The segmentation of chromosome images using this algorithm is compared with traditional methods. The results show that the algorithm in this paper has complete segmentation of chromosome images with the smallest error. The algorithm in this paper can achieve an accurate segmentation of chromosome images.