摘要
图像分割和对象提取是从图像处理到图像分析的关键步骤。K-均值聚类算法和粒子群优化方法结合,即将K-均值方法的结果作为一个粒子并采用粒子群优化的方法,通过适应度函数,利用新的分类中心调整粒子位置,产生新的聚类中心。并将此方法应用于图像的分割。最后,将两种方法的处理结果进行了比较,结果表示基于PSO聚类方法对图像的分割效果比原算法有所改进。
Image segmentation and object extraction are the key steps in image processing. We put forward a new algorithm combining Particle Swarm Optimization （PSO） with K-means algorithm. In the new algorithm, the results of K-means algorithm is used as a particle, and PSO is used to generate a new clustering center by adjusting the particle position with new classification center through fitness function. The algorithm is used in image segmentation to get the center of clustering. Results of K-means algorithm and PSO based clustering method were analyzed, which showed that the later fits the image segmentation better.
出处
《电光与控制》
北大核心
2009年第2期5-6,17,共3页
Electronics Optics & Control
基金
湖南省创新基金(06C26214301140)
关键词
粒子群优化
聚类算法
K-均值聚类
图像分割
桢式识别
图像处理
particle swarm optimization
clustering method
K-means clustering
image segmentation
pattern recognition
image processing
