摘要
针对包含有OCR字符区域的OCR证件图象,提出了一种基于方向投影和类内散度的OCR证件图象的倾斜检测算法.该算法以整数Bresenham算法计算不同扫描角度的方向投影,同时计算各个方向投影的最小总类内散度,然后寻找所有最小总类内散度的最小值,以该最小值对应的扫描角度为最终的检测倾角.实验结果表明,该算法能快速准确地检测出OCR证件图象的倾斜角度.
According to the OCR ID Doctment image containing the OCR character region, a new skew detection method based on the directional projection and the within-group measure of dispersion has been presented. Several directional projection are calculated in the method, using Bresenham＇s integer line drawing algorithm. After calculating minimal within-group measure of dispersion of the directional projection and searching minimal value of all within-group measure of dispersion, the corresponding angle of minimal value is the tilted angle of the OCR ID document. Experiment results show that the slanting angle of a OCR ID document image can be detected accurately.
出处
《广西工学院学报》
CAS
2009年第2期38-41,46,共5页
Journal of Guangxi University of Technology
基金
福建省科技厅重点项目(2008H0016)资助
广西教育厅科研项目(200707LX196
200707LX182)资助
关键词
OCR证件
方向投影
类内散度
倾斜检测
OCR ID document
directional projection
within-group measure of dispersion
skew detection
