摘要
针对细菌觅食算法中细菌趋化方向选择的随机性,以及趋化操作中更新细菌位置没有参照菌群群体最优位置以及个体历史最优位置信息等问题,提出改进的细菌觅食算法(IBFA)。首先,细菌趋化方向的选择,结合细菌自身学习及群体合作,避免浪费细菌到达最优解的时间;其次,粒子群算法应用到趋化步骤中,更新细菌位置参照菌群群体最优位置以及个体历史最优位置信息,以提高算法全局搜索能力和收敛速度;最后,由IBFA算法优化最大类间方差函数来取得最优阈值。IBFA算法的性能分别与传统最大类间方差法、基于细菌觅食的多阈值图像分割结果对比,实验结果表明,本文算法能有效缩短图像分割时间,提高图像分割准确率。
Because the selection of bacteria chemotaxis direction has the randomness in bacteria foraging algorithm （BFA） and bacterial updates positions without referring to the optimal location of bacterial colony and the optimal location information of individual history, so the improved bacteria foraging algorithm is presented （IBFA）. First, the strategy of the bacterial chemotaxis direction is combined with the bacteria individual cognitive and group cooperation, so it can avoid wasting bacteria to the optimal solution time;Second, parti- cle swarm optimization （PSO） algorithm is applied to the chemotactic step, bacterial updates positions referring to the optimal location of bacterial colony and the optimal location information of individual history, to improve the global search ability and convergence speed ; Finally, maximum between - class variance method is optimized by IBFA algorithm to obtain the optimal thresholds. The performance of IBFA algorithm is compared with Otsu and the multi -threshold image segmentation based on BFA algorithm, the experimental results show that the IBFA algorithm can effectively shorten the time of image segmentation and improve the accuracy of image segmentation.
作者
田云
车亚琴
薛一兰
TIAN Yun;CHE Ya - qin;XUE Yi - lan(Xinzhou Teachers University, Xinzhou 034000, Chin)
出处
《忻州师范学院学报》
2018年第2期49-53,84,共6页
Journal of Xinzhou Teachers University
基金
山西省自然科学基金项目(2014011019-3)
山西省重点实验室开放课题基金项目(2016002)
学院重点学科项目(XK201404)
院级青年基金项目(QN201409)
关键词
细菌觅食算法
粒子群算法
多阈值图像分割
bacterial foraging algorithm
particle swarm algorithm
muhilevel thresholding image segmentation
作者简介
田云（1982-），女，河北保定人，忻州师范学院计算机系讲师，硕士，从事智能计算、图像处理研究。
