摘要
针对Yanowitz和Bruckstein的二值文字图像后处理算法存在的问题,提出了一种新的基于灰度期望值的后处理算法.实验结果表明,该算法能自动选取阈值、运算速度快、处理效果好,具有良好的噪声适应性.
An improved postprocessing approach based on gray level averaging is described. The proposed method can effectively solve the problems existing in Yanowitz and Bruckstein's postprocessing method. Experimental results show that both the quality of segmentation and computation speed are improved compared to Yanowitz and Bruckstein's method.
出处
《上海大学学报(自然科学版)》
CAS
CSCD
2003年第4期283-286,共4页
Journal of Shanghai University:Natural Science Edition
基金
上海市教委发展基金 (2 0 0 0 A45)资助项目
关键词
文字图像
二值化
后处理
document image
binarization
post-processing