提出了一种改进的加权Hausdorff距离,并将其应用于字符图像的匹配。该方法根据字符图像的结构特征对字符不同区域设置不同的权重。实践表明该方法改善了图像匹配效果。
An improved approach based on weighted Hausdorff distance with application to character image matching was proposed. Thanks to the idea that different feature should correspond to different weight, different weights were set for different regions of a character image in the proposed method. Experimental results showed that the improved method is promising and outperforms the original weighed Hausdorff distance.