基于广义混沌混合PSO的快速红外图像分割算法

General Hybridized PSO with Chaos for Fast Infrared Image Segmentation Method

为了准确的实现红外目标识别,提出了一种基于广义混沌混合PSO的快速红外图像分割算法.二维模糊划分最大熵分割方法不仅利用了灰度信息以及空间邻域信息,而且兼顾了图像自身的模糊性,能取得较为满意的分割结果.该方法实质上是一种具有搜索空间大、多局部极值点的典型非线性整数规划问题.广义混沌混合PSO算法在广义PSO算法的基础上,引入自适应平衡搜索,当算法发生停滞时引入模拟退火机制有选择地对当前全局最优粒子进行混沌优化,在增强局部搜索能力的同时能够克服早熟收敛现象.实验证明,运用广义混沌混合PSO算法实现红外图像二维模糊划分最大熵分割是快速、稳定的.

To detect infrared objects accurately, a fast infrared image segmentation method based on general hybridized PSO with chaos is proposed. The method of 2-D maximum fuzzy partition entropy can obtain better segmentation,because it takes advantage of gray and spatial neighboring information, and fuzziness of image also is taken into consideration. In essence,it is a typical nonlinear integer programming problem with huge searching space and many local optima. General hybridized PSO with chaos is based on general PSO, and it makes use of adaptive balance searching strategy. When the evolution stops, simulated annealing algorithm is introduced to select the current global optimum to be chaotic optimized for the sake of enhancing local searching ability and overcoming premature convergence. Experimental results show that the method can segment infrared image quickly and stably.