基于图像OMR的预处理技术研究

Research of Pre-processing Algorithm Based on Image OMR

阐述了信息填涂卡文档图像标记识别系统中的预处理算法,并提出一种新的信息卡填写方式。对信息填涂卡文档图像预处理中的利用计算机图像处理技术将彩色图像转化成灰度图像、二值化以及图像平滑等关键步骤进行了描述。在图像二值化过程中采用全局阈值法,在图像平滑过程中采用改进的邻域平均算法。结果表明这些算法能有效地解决信息填涂卡文档图像预处理问题。

Discussed a series of pre- processing theory and algorithms for analyzing document image of information card based on optical mark recognition,and proposed a new information card fill - in method. The major steps of the computer image processing techniques were also described, which include： transforming of the color plate to gray plate, image binarizing, image smoothing, etc. Global threshold method was applied to image hinarizing. An improved neighborhood averaging method was applied to image smoothing. The result showed that these methods could meet the need of document image pre - processing.