摘要
在具有噪声及灰度不均匀的复杂背景中提取出待识别的字符， 是字符识别的关键一步。提出应用简单统计法及Ｒｏｂｅｒｔｓ边缘检测相结合的方法， 对图像进行二值化处理， 并在此基础上利用Ｈ－Ｓ连同分析方法去掉了复杂的背景块。在５８６／ＡＭＤＫ６／２００机器上， 对５０余幅图像处理， 均正确提取出了待识别的字符且每幅的实现时间不到１ ｓ。
Extracting characters to be recognized from the complex background with noises and uneven gray levels is a key step to recognize characters. This paper presents a method of combining the simple statistic method with Roberts edge detection to carry out the image binary, on the basis of which the method of analysis of H S connectivity is used to remove the complex background block. The recognized characters are extracted right by processing fifty images on 586/AMDK 6/200 computers and the executing time is less than one second.
出处
《西安理工大学学报》
CAS
1999年第3期50-53,共4页
Journal of Xi'an University of Technology
