摘要
在有些字符识别的应用领域,待识别的字符位于圆环区域内,例如圆形工业产品上的编号等,这给识别带来不便。因此先要把圆环区域转换成矩形区域,调整字符位置呈水平排列后再进行字符识别。采用HALCON机器视觉软件进行的设计,主要步骤包含圆环区域的极坐标转换算法和字符识别算法的实现。实验表明,核心算法有效可行,字符识别率高。
In some character recognition applications, the characters are distributed in annular region such as serial numbers on circular industrial products, which are not easy to recognize. So it must turn the annular region which contains the characters into rectangular region first and do the character recognition later. The design based on HALCON includes the implementation of region polar transformation and character recognition. The experiment shows that the key algorithms are effective and feasible with high accuracy.
出处
《现代计算机》
2010年第7期58-60,71,共4页
Modern Computer