二维灰度图像的分形维数计算

Estimation of fractal dimension for 2-D gray image

分析了当前计算图像分形维数的算法普遍存在对高维数图像计算误差较大且计算量大的缺点。结合计盒维数算法的原理,给出一种新的盒子边界。理论和实验证明,该算法较Chaudhuri和Sarkar给出的经典边界选取在准确性和实时性上都有所改进。

The general fractal dimension algorithms of image have the disadvantage of the precision and the speed especially when applied to the high dimension images for the moment. Based on the box counting method, a new bound of box size was put forward. Experiment and theoretical results show that the algorithm is better than Chaudhuri and Sarkar＇s classical bound both on the precision and the speed.