一种新的图像阈值分割算法

提出了一种基于区域灰度均值-梯度共生矩阵模型和二维Renyi熵原理的阈值分割方法.该方法不仅利用了图像的灰度信息,而且也利用了梯度信息,通过计算基于灰度-梯度共生矩阵的二维Renyi熵,并使边缘区域的熵最大来选择阈值.二维Renyi熵比其它二维熵多了一个参数,此参数使得二维Renyi熵能处理更多类型的图像.将上述方法进行仿真实验,并且仿真其它二维熵方法,仿真结果显示,该算法比其他二维熵方法效果更有效更一般化.

A New Algorithm of Image Threshold Segmentation

A threshold segmentation algorithm based on commensal matrix of region gray average value-grad model and two-dimensional Renyi's entropy is presented.The algorithm uses not only gray information but also grad information in image.The threshold value is gained by calculating two-dimensional Renyi's entropy based on commensal matrix of region gray average value-grad.Two-dimensional Renyi's entropy has an additional parameter comparing to other two-dimensional entropy.The additional parameter makes it process more types of image.The simulation result shows that this algorithm is better than other two-dimensional entropy algorithms.