Abstract:

The general method of image segmentation has always been a hot and difficult point in image processing. With the rise and development of artificial intelligence, swarm intelligence algorithms have become the current hot research direction. Image segmentation technology combined with swarm intelligence algorithms has become a new and effective improvement method. The swarm intelligence algorithm simulates the law of action of things in nature, combines artificial intelligence and swarm biology, searches for the optimal solution in the solution space, and provides new solutions to solve complex problems. It describes the research status and development process of swarm intelligence algorithms, combining the early Ant Colony Optimization（ACO）, the classic Particle Swarm Optimization Algorithm（PSO） and the newer Sparrow Search Algorithm（SSA） as an example to introduce its algorithm principles and methods in detail, and briefly describes the principles of Bat Algorithm（BA）, Whale Optimization Algorithm（WOA）, Artifical Bee Colony Algorithm（ABC）, Firefly Algorithm（FA）, Cuckoo Search（CS）, Bacterial Foraging Optimization（BFO） and the latest Mayfly Algorithm（MA）, based on this, combines with domestic and foreign literature to analyze and summarize the improved method of the above algorithm and the comprehensive improvement and application of image segmentation technology. The swarm intelligence algorithm combined with the representative algorithms of image segmentation technology is extracted for list analysis and summary, and then the unified framework, common characteristics, different differences and existing problems of the swarm intelligence algorithm are summarized. Finally, the future trend is prospected.

Key words: swarm intelligence algorithm, image segmentation, ant colony algorithm, particle swarm optimization algorithm, sparrow search algorithm, bat algorithm, whale optimization algorithm, mayfly algorithm

摘要：

图像分割的通用方法一直是图像处理领域中的热点和难点。随着人工智能的兴起和发展，群体智能算法成为当下热点研究的方向，将图像分割技术结合群体智能算法成为一种新型有效的改进方法。群智能算法通过模拟自然界的事物或生物的行动规律，将传统的人工智能和群体生物结合，在解空间中搜索最优解，为解决复杂问题提供了新的解决思路。阐述群体智能算法的研究现状和发展过程，将早期的蚁群算法（Ant Colony Optimization，ACO）、经典的粒子群算法（Particle Swarm Optimization Algorithm，PSO）以及较新的麻雀搜索算法（Sparrow Search Algorithm，SSA）为例详细介绍其算法原理方法，并简要表述蝙蝠算法（Bat Algorithm，BA）、鲸鱼优化算法（Whale Optimization Algorithm，WOA）、人工蜂群算法（Artificial Bee Colony Algorithm，ABC）、萤火虫算法（Firefly Algorithm，FA）、布谷鸟搜索法（Cuckoo Search，CS）、细菌觅食算法（Bacterial Foraging Optimization，BFO）和最新的蜉蝣算法（Mayfly Algorithm，MA）的原理，在此基础上，结合国内外文献对上述算法的改进方法和结合图像分割技术的综合改进及应用进行分析总结。将群体智能算法结合图像分割技术的代表性算法提取出来进行列表分析总结，随后概述总结群体智能算法的统一框架、共同特性、不同的差异并提出存在的问题，最后对未来趋势做出展望。

关键词: 群体智能算法, 图像分割, 蚁群算法, 粒子群算法, 麻雀搜索算法, 蝙蝠算法, 鲸鱼优化算法, 蜉蝣算法