关键词: 遗传算法 熵 骨髓细胞 分割 种群

Abstract:

This paper proposes a genetic clustering image segmentation algorithm based on entropy. Better segmentation effect can be attained by taking the gray   levels of pixels as the eigenvector, taking advantage of the histogram entropy principle function as the fitness function, adopting the ranking selectio n operation, making use of arithmetic crossover and mutation at a certain probability, and combining with the clustering analysis to initialize the clus tering center of the population to segment cell images with genetic algorithms.

Key words: （GA, entropy, marrow cell, segmentation, population）