改进的二维阈值图像分割

Improved Two-dimensional Image Threshold Segmentation

本文提出了一种基于灰度图像像素点灰度和点邻域方差的改进二维图像分割法。新方法改进了阈值判定域,考虑了边界和噪声的影响,新定义了一个阈值分割函数,提高了分割精度。利用思维进化算法优化分割参数,提高了最优阈值的寻找速度。实验结果表明,基于思维进化算法的改进二维图像分割法优于传统算法,该算法具有较好的稳定性和收敛速度,更能满足图像处理高效率、短时耗的要求。

This paper proposes a improved two-dimensional image segmentation algorithm based on pixel gray level and pixel neighborhood variance of the gray-scale image.This new algorithm improves the judgment domain of threshold,considers the influence of the boundary and noise,defines a new threshold segmentation function,increases the accuracy of segmentation.Using mind evolutionary algorithm to optimize the segmentation parameters,and the search speed of optimal threshold is increased.The experiment result proves that the proposed algorithm is better than traditional algorithm,it has better stability and convergence speed,and also can meet the requirements of high efficiency,short-term consumption in the image processing.