基于灰度级-平均灰度级直方图的现有二维交叉熵阈值分割法的分割结果不够准确，计算最佳阈值时需搜索整个解空间，因而效率不高. 针对这一问题，提出一种基于灰度-梯度共生矩阵的二维最小交叉熵阈值选取快速迭代算法，推导了相关的公式. 对典型测试图像进行了大量实验，并与基于灰度级-平均灰度级直方图的方法在分割结果及运行时间上作了比较，结果表明所提出的算法分割结果更加精确，且计算最佳阈值时只需遍历其中一小部分解空间，运行时间减少到5%左右.

The existing two-dimensional cross entropy thresholding method based on gray level-average gray level histogram does not produce accurate enough results. In addition, it needs to search the entire solution space to obtain the best threshold. In this paper, a fast iterative algorithm based on gray level-gradient cooccurrence matrix is proposed for selecting two-dimensional minimum cross-entropy threshold. The algorithmic formulas are derived. Experiments are carried out on typical test images. Comparisons of segmentation results and execution speed are made between the proposed method and the method based on gray level-average gray level 2D histogram. The results show that the proposed algorithm provides better segmentation. Only a small part of the solution space needs to be searched to find the best threshold. The running time reduces to about 5% of the method based on gray level-average gray level 2D histogram.