基于自适应粒子群算法和数据场的图像二维阈值分割

针对图像分割中最优阈值选择的问题,将粒子群优化算法和数据场理论相结合,提出一种图像二维阈值分割算法.首先把数据场的理论引入到图像处理中,将图像的灰度值空间映射到数据场的势空间;然后通过自适应的粒子群优化算法寻找数据场中最大势值,该势值对应最优阈值;最后根据找到的阈值进行图像分割.在进行空间映射的过程中,将二维直方图中的序偶?p,q?视作数据对象,其中p代表像素的灰度值,q代表邻域的灰度值,选用拟核力场高斯势函数计算各数据对象之间的相互作用,生成了二维直方图的三维数据场.文中亦对数据场的各个参数进行了详尽的探讨.实验结果表明,文中算法不仅合理、有效,而且大大降低了计算的复杂性,能够适应大多数图像的分割.

Image Segmentation with Two-Dimension Threshold Based on Adaptive Particle Swarm Optimization and Data Field

A novel method of image segmentation based on adaptive particle swarm optimization and data field has been proposed for optimal threshold selection in image segmentation.In the proposed method,images are mapped from the grayscale space to the potential space of the data field.By taking the frequency of two-dimension gray histogram as the mass of data field,the interactions between elements in the two-dimension histogram can be calculated,a three-dimension data field is generated subsequently.Thus,by employing adaptive particle swarm optimization,the optimal threshold,which is the point with the maximum potential value,can be found and good segmentation results can be obtained.The relevant experiments have shown that the proposed method is effective and greatly reduces the complexity of computation.