摘要
本文主要在模式识别领域,研究人工打码字符图像在图像预处理之后,识别之前的单个整体字符的倾斜矫正算法。对于实际的字符图像源,由于是人工逐个字符进行打码,难免会出现字符倾斜的状况,再加上工件背景比较复杂,拍摄光照不均等原因,使得对字符图像的处理和倾斜矫正愈加困难。本文结合实际应用,在对字符图像进行前期的预处理和字符分割之后,再对分割后的单个字符进行倾斜矫正,实验取得了良好的效果,为字符后期的识别创造了有利的条件。
In the area of pattern recognition, a tilt correction algorithm for artificial coding character image which is been preprocessed before recognition is researched and presented in this paper. Because of artificial coding, tilting character is unavoidable. And due to the complicated workpiece background and uneven lighting, the processing and tilt correction for character image become much more difficult. Combining with practical application in this paper, tilt correction for single character is done after image pre-processing and character segmentation of character image. The experiment shows that the algorithm is effective, and creates a favorable condition for the subsequent character recognition.
出处
《软件》
2012年第12期139-141,共3页
Software
关键词
模式识别
倾斜矫正
打码字符
图像预处理
Pattern Recognition
Tilt Correction
Coding Character
Image Pre-processing
