基于改进PSO理论的二维Otsu分割算法

A Two-Dimensional Otsu Segmentation Algorithm Based on Improved Particle Swarm Optimization

二维Otsu阈值分割方法是图像分割中常用的分割方法,但其算法运算量大,大大限制了实时性要求,为此,提出了一种基于改进PSO理论的二维Otsu分割算法。首先从二维Otsu算法基本理论出发应用有智能寻优特点的PSO理论来寻找最佳阈值向量,而后又针对传统PSO算法的缺点引入协同分工的合作思想来寻找最佳的分割阈值。实验证明,应用改进的PSO理论的二维Otsu算法不仅能够正确地寻找到阈值,还大大提高了计算速度,是一种高效快速的分割方法。

The two-dimensional Otsu segmentation algorithm is a commonly used segmentation method, which spends so much time in the segmentation that its real-time performance is limited. Therefore, a two- dimensional Otsu segmentation algorithm based on improved particle swarm optimization was proposed. First, the PSO theory that has intelligent optimizing ability was used for seeking the best threshold value vector based on the two-dimensional Otsu algorithm fundamental theory. Then, aiming at the shortcoming of the traditional PSO algorithm, the coordination and responsibility assignment thought was introduced to seek for the best division threshold value. Simulation result indicated that this algorithm can not only seek the correct threshold, hut also improve the calculation speed, which is an efficient segmentation method.