- 90
- 0
- 约7.97千字
- 约 14页
- 2018-10-16 发布于福建
- 举报
工业ECC200码快速定位算法的研究
工业ECC200码快速定位算法的研究
摘 要:二维码优点众多,被广泛应用于各行业特别是工业生产领域,但由于工业环境差、工业生产产品有效二维码比例小、产品材质导致的二维码打印显示图案不同等原因,采集到的二维码图像存在光照不均或无法检测等情况,阻碍二维码优点的有效发挥。鉴于此,以工业ECC200码为研究对象,首先采用多尺度Retinex算法对因光照因素造成的异常图像进行处理,然后采用图像平面漫游的方式对复杂背景下的ECC200码进行定位,对圆点式ECC200码使用模板匹配的方式进行精确定位,从而提高识别准确率和识别时间。与常见二维码识别软件进行对比实验表明,提出的方法在识别时间和识别率上都更胜一筹。
关键词:工业ECC200码;多尺度Retinex;图像平面漫游;模板匹配搜索
DOI:10.11907/rjdk.171520
中图分类号:TP312 文献标识码:A 文章编号:1672-7800(2017)009-0061-04
Abstract:The many advantages of two-dimensional code is the main reason why it is widely used in various industries, especially in industrial production.However, due to poor industrial environment, industrial products effective two-dimensional code is small,the different two-dimensional display caused by the product material, the two-dimensional code image is uneven illumination, or can not detect ,which hinder the advantages of the effective play of two-dimensional code. In this paper, the industrial ECC200 code is used as the study object. Firstly, the multi-scale Retinex algorithm is used to deal with the abnormal image caused by the illumination factor. Then, the ECC200 code in the complex background is located by image plane roaming. The dotted ECC200 codes are located using template matching to improve recognition accuracy and recognition time. Compared with the common two-dimensional code recognition softwares, it is shown that the method proposed in this paper is superior to recognition time and recognition rate.
Key Words:industrial ECC200 code; multi-Scale Retinex; image plane roaming; template matching search
0 引言
相较于一维码,二维条码具有信息量大、可靠性高、保密性好等优点[1]。近年来,二维条码技术广泛??用于工业控制领域,很多工业零部件、电子芯片和产品包装均采用二维条码技术进行标识,完成质量、产品溯源等功能[2]。较常用的一种是矩阵式二维条码(Data Matrix,以下简称DM码),Data Matrix码又可分为ECC000-140与ECC200码两种。其中, ECC200码因其容易演算、信息容量大及容错率高等优点,被广泛应用于工业领域。
依据ECC200码的图案特征,邹沿新等[3]提出根据码型特征搜索“L”型寻边区,实现了ECC200码的自动识别算法,可以克服背景噪声、几何失真等影响,但如果条码背景过于复杂,感兴趣区域难以精确会发生误检现象。王伟等[4]提出通过峰度值排序法及模块区域微调法进行条码定位,再利用遗传算法提取二维码信息的算法,其优点是鲁棒性好、可靠性高,缺点是处理速度慢,很难实现实时应用。工业ECC
原创力文档

文档评论(0)