基于Lab颜色空间的融合改进二进制量子PSO和Otsu优化算法  

IMPROVED BINARY QUANTUM PSO AND OTSU OPTIMIZATION ALGORITHM BASED ON LAB COLOR SPACE

作　　者：徐武[1] 文聪 唐文权 郭兴 Xu Wu;Wen Cong;Tang Wenquan;Guo Xing(Institute of Electrical and Information Engineering,Yunnan Minzu University,Kunming 650500,Yunnan,China)

机构地区：[1]云南民族大学电气信息工程学院,云南昆明650500

出　　处：《计算机应用与软件》2022年第6期265-268,349,共5页Computer Applications and Software

基　　金：国家自然科学基金项目(U1802271)。

摘　　要：针对通过RGB空间各个颜色分量高度线性相关会导致分割过程处理速度慢、分割结果不准确的问题,以及二维Otsu算法实时性差,计算量大的缺陷,提出一种基于Lab空间的融合改进二进制量子PSO和Otsu优化算法。将RGB空间转换为Lab空间,减弱不同颜色分量的相关性;利用改进的二进制量子粒子群算法搜索最优阈值,作为改进二维Otsu算法的分割阈值,然后对图像进行分割。通过实验证明改进算法与其他算法相比,分割速度和分割精确度都有很大的提升。The highly linear correlation of color components in RGB space causes the low processing speed and inaccurate segmentation results,and the two-dimensional Otsu algorithm has poor real-time performance and costs large amount of calculation.To solve these problems,we propose an improved binary quantum PSO and Otsu optimization algorithm based on lab space.RGB space was transformed into lab space to reduce the correlation of different color components.We used the improved binary quantum particle swarm optimization to search the optimal threshold as the segmentation threshold of the improved 2D Otsu algorithm.The image was segmented.The experiments show that compared with other algorithms,the improved algorithm greatly improves the segmentation speed and accuracy.

关 键 词：二维OTSU Lab颜色空间 二进制量子PSO 语义分割 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]