摘要： 提出了一种快速检测CCD图像中条形码区域的方法。首先等间隔抽取扫描行逐点平滑和梯度化,其次用动态全局阈值对目标行二值化,进一步搜索特征点并做筛选和分类,最后用最小二乘法拟合边缘,获得目标区域的特征参数提取目标区域同时进行倾斜校正。等间隔抽取扫描行的处理办法成倍的减小了处理的数据量。对取自30种商品不同倾角的84幅图像提取结果表明当倾角在±35o以内时正确率为100%。

关键词: 条形码, 动态全局阈值, 特征点, 区域提取

Abstract: In this paper a method of bar code region quick detection applied to bar code CCD image is presented. Firstly, rows extracted from original image every several rows are smoothed and gradient of pixels is calculated. Secondly, binary image composed of target rows is got by the dynamic global threshold. Then, character points sought are filtrated and labeled. Finally, with method of least square fit, they are fitted to form edges of object region. Object region is extracted by its character parameters and its skew is corrected at the same time. The method of row extracting reduces data needed to process several times. Experiment result of 84 images got from 30 kind of commodity indicates that when skew is within ±35o the result is correct entirely.

中图分类号: 

• TP391.44