混沌映射麻雀搜索优化OTSU的图像分割算法

打开文本图片集

中图分类号： TP391.4 文献标志码：A

Abstract： To address the time-consuming and overly subjective challenges of lesion segmentation in dermatoscope images， an improved sparrow optimization algorithm （ISSA） is proposed to optimize the OTSU threshold segmentation. The algorithm， inspired by the foraging and antipredation behavior of sparrows，utilizes the inter-class variance as the fitness function.By incorporating the piecewise linear chaotic map （PWLCM） in population initialization， it enhances the algorithm's search space and optimization performance， facilitating the escape from local optima in a timely manner. The proposed ISSA is compared with commonly used optimization algorithms， including the particle swarm optimizer （PSO）， grey wolf optimizer （GWO）， and the original sparrow search algorithm （SSA）， in dual-threshold OTSU segmentation experimentson dermatoscope images.The results demonstrate that ISSA not only shows an improvement in optimization but also reduces the number of iterations by 92.2% 68.2% ，and 41.7% and the runtime by 66.4% 0 43.4% ，and 21.1% compared to PSO， GWO， and SSA， respectively.

Keywords： image segmentation; sparrow search algorithm; OTSU algorithm; PWLCM chaotic mapping; dermatoscope image

引言

色素性皮肤病是由于黑素细胞和黑素生成异常造成，表现为色素增多或减少引起了皮肤颜色变化的一种常见皮肤病，可由遗传及环境因素引起。（剩余9073字）