一种改进樽海鞘群算法优化K-Means的小麦覆盖度提取方法  

Wheat coverage extraction based on improved salp swarm algorithm for optimizing K-Means

作　　者：王向[1] 李月凤 王震洲[1] 张佳佳[2] WANG Xiang;LI Yuefeng;WANG Zhenzhou;ZHANG Jiajia(School of Information Science and Engineering,Hebei University of Science and Technology,Shijiazhuang,Hebei 050018,China;Meteorological Detection Technology Section,Hebei Meteorological Technology and Equipment Center,Shijiazhuang,Hebei 050021,China)

机构地区：[1]河北科技大学信息科学与工程学院,河北石家庄050018 [2]河北省气象技术装备中心气象探测技术科,河北石家庄050021

出　　处：《河北科技大学学报》2023年第4期356-367,共12页Journal of Hebei University of Science and Technology

基　　金：国家自然科学基金(2021YFC0863200-6)。

摘　　要：针对K-Means算法对初始聚类中心的依赖性较高,容易出现局部最优停滞的问题,提出一种改进樽海鞘群算法优化K-Means的小麦覆盖度提取算法。首先,将小麦图像转换到HSV色彩空间;然后,用改进樽海鞘群算法进行全局寻优,以获得全局最优值作为K-Means算法的初始聚类中心,接着运用K-Means算法进行局部寻优,直到迭代完成;最终,输出经过分割的小麦图像。为了评估算法性能,使用12个基准函数对ISSA及其他智能优化算法进行对比测试,同时将改进樽海鞘群算法优化K-Means应用于小麦覆盖度提取。结果表明,ISSA算法在优化精度和收敛速度上均超越其他算法,鲁棒性也得到了显著提高。与其他算法相比,ISSA-K算法分割后的小麦图像纹理比较清晰,效果更佳,同时具有更加高效的优势,可用于小麦覆盖度的提取,具有较强的实用性。Aiming at the problems of high dependence of the K-Means algorithm on the initial clustering center and local optimal stagnation,a wheat coverage extraction algorithm with optimized K-Means by an improved salp swarm algorithm was proposed.First,the wheat image was converted to HSV colour space;Then the improved salp swarm algorithm was used to find the global optimal value as the initial clustering center of K-Means algorithm;Afterwards,the K-Means algorithm was used for local optimization until the iteration was completed;Finally the segmented wheat image was output.In order to evaluate algorithm performance,ISSA and other intelligent optimization algorithms were compared and tested by using 12 benchmark functions.Meanwhile,the improved salp swarm optimization K-Means algorithm was applied to wheat coverage extraction.The results indicate that the optimization accuracy and convergence speed of ISSA algorithm are superior to other algorithms,and its robustness is also significantly improved.Compared with other algorithms,the wheat image segmentation by ISSA-K algorithm has clearer texture and better effect.At the same time,it has the advantage of being more efficient which can be used for wheat coverage extraction and has strong practicability.

关 键 词：图像处理 K-MEANS 改进樽海鞘群算法 HSV色彩空间 图像分割 小麦覆盖度提取 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]