激光与光电子学进展, 2015, 52 (5): 051001, 网络出版: 2015-05-13
结合Predator-Prey-AACO的图像边缘检测算法 下载: 512次
Image Edge Extraction Combined with Predator-Prey-AACO Algorithm
图像处理 边缘检测 自适应蚁群优化算法 生物捕食-被捕食行为 自适应调整策略 image processing edge detection adaptive ant colony optimization algorithm biological predator-prey behavior adaptive strategy
摘要
针对自适应蚁群优化(AACO)算法在图像边缘提取中经常出现效率低、易陷入局部极值等问题,提出一种结合生物Predator-Prey 行为的自适应蚁群图像边缘检测算法。该算法将Predator-Prey 行为与AACO 算法相结合,将蚁群分成Predator 种群和Prey 种群,初始阶段利用AACO 算法进行搜索两种群,一定迭代次数后,两种群进入排斥阶段;通过自动阈值法提取图像边缘。实验结果表明,与AACO 算法和Canny 算法相比,在精确度方面,该算法提取的图像边缘明显优于前两种算法提取的边缘;同时保持了AACO 算法收敛速度快的特点,并克服了其易陷入局部极值等缺点;因此,该算法能够高效准确地检测出图像边缘。
Abstract
In view of adaptive ant colony optimization (AACO) algorithm in image edge extraction often appears problems of low efficiency, and easily falling into the local extremum. An adaptive ant colony image edge detection algorithm combined with biological Predator-Prey behavior is proposed. The algorithm is combined with AACO algorithm and Predator-Prey behavior. Ant colony is divided into Predator and Prey species, AACO algorithm is used to search the two species in the initial stage, after a certain number of iterations, the two species enter the exclusion stage. The method of automatic threshold is used to extract image edge. Experimental results show that compared with AACO algorithm and Canny algorithm, in terms of accuracy, image edge extrated by the proposed algorithm is better than the edge extracted by the first two algoritms. Meanwhile, it maintains fast convergence characteristic of AACO algorithm, and overcomes its shortcoming of easily falling into local extremum. So image edge can be effectively detected by it.
惠晓威, 常正英, 林森, 曹益华. 结合Predator-Prey-AACO的图像边缘检测算法[J]. 激光与光电子学进展, 2015, 52(5): 051001. Hui Xiaowei, Chang Zhengying, Lin Sen, Cao Yihua. Image Edge Extraction Combined with Predator-Prey-AACO Algorithm[J]. Laser & Optoelectronics Progress, 2015, 52(5): 051001.