摘要
为实现计算机对文本图像的自动分类和鉴别,对文本图像分割进行研究。建立基于贝叶斯决策的最小错误率文字分割模型。利用最小二乘法估计模型参数最后在多颜色空间中选择最优分割阈值,取得了理想的图像分割效果。
For documental image recognition, retrieving and identification automatically, image segment is analyzed. Based on Bayes segment model is established with minimum error risk, and parameters are estimated by least squares method. At last, optimization threshold is obtained by multi-color space analyzing and promising results are obtained.
出处
《科学技术与工程》
2009年第16期4826-4827,共2页
Science Technology and Engineering
基金
湖北省科技攻关计划项目(2006AA301B44)资助
关键词
图像分割
文本处理
贝叶斯决策
image segment document processing Bayes decision
