边缘检测算法在污损条码识别中的应用研究  

Application Research of Edge Detection Algorithm in Defected Barcode Recognition

作　　者：胡大辉[1] 

机构地区：[1]西南大学,重庆402460

出　　处：《包装工程》2013年第3期104-108,共5页Packaging Engineering

基　　金：国家自然科学基金资助项目(21007051);西南大学青年基金资助项目(2010RCQ003)

摘　　要：针对污损条码识别速度慢和正确识别率低的问题,提出了利用图像边缘检测算子来提取污损条码的边缘特征并进行识别。分析了常用边缘检测算法的算子特征后,采用Sobel算子进行了条码的边缘检测。针对原算子的不足,提出了一种改进的Sobel算子,改进算子增加2个方向的运算模版,能克服原算子在边缘提取时图像线条较粗等缺陷。仿真实验表明,改进的算子能快速有效地识别污损条码,正确率高达98.5%。A algorithm using image edge detection operators to extract the edge of defaced barcode and to identify it was put forward to solve the problems of slow identification speed and low recognition rate. Sobel operator was applied to detect the barcode edge after analysis of the characteristics of existing commonly used edge detection operators. An improved Sobel operator was put forward, which increase two directions template, which can overcome the thicker lines and other defects of the original operator in image edge detection. The experiments showed that the new operator can quickly and efficiently identify defaced barcode; the correct rate is 98.5%.

关 键 词：边缘检测 污损条码 SOBEL算子 边缘细化 

分 类 号：TS851.6[轻工技术与工程] TS896