APP下载

基于图像处理的化工产品英文字符识别研究

2021-06-29王蕊

粘接 2021年5期

王蕊

摘 要:针对传统板材喷码人工记录的问题,提出一种基于图像处理的喷码图像识别方法。在该方法中,首先采用加权平均法对图像进行灰度化处理,然后采用背景差分法和Otsu对图像进行定位和分割;利用旋转投影法和最小二乘拟合法对喷码图像进行校正;通过PCA融合方法完成对一维投影特征、外轮廓特征和网格特征的提取;最后通过二叉树的SVM进行识别分类。通过试验验证表明,本文采用的方法可有效识别不同格式的英文字符,且识别正确数明显高于传统的SVM方法,具有一定的优势。

关键词:喷码图像;英文字符;图像识别;SVM分类

中图分类号:TP391.43 文献标识码:A 文章编号:1001-5922(2021)05-0076-04

Research on English Character Recognition of Chemical Products Based on Image Processing

Wang Rui

(Xi an Siyuan University,Xi an 710038,China)

Abstract:Aiming at the problem of manual record of traditional plate spray code, an image recognition method based on image processing is proposed. In this method, firstly, the weighted average method is used to gray the image, then the background difference method and Otsu are used to locate and segment the image; the rotation projection method and the least square fitting method are used to correct the spray code image; the PCA fusion method is used to extract the one-dimensional projection features, the contour features and the grid features; finally, recognition and classification are carried out through the SVM of the binary tree. The experimental results show that the proposed method can effectively recognize English characters of different formats, and the correct number of recognition is significantly higher than that of the traditional SVM method , which has certain advantages.

Key words:inkjet image; English characters; image recognition; SVM classification

为区别钢铁板材信息,部分厂家会在生产的板材表面会通过喷涂的形式喷印上板材的信息,如钢板号、钢种、尺寸、制造商、生产批次等,且每一个板材产品对应于唯一的钢板号。在这其中,包含着大量的英文字符信息。而板材成品从入库导装车发货过程中,管理员往往会通过手工记录的方式来记录本批次板材的喷码信息,以此区别不同类别的板材。但这种传统的人工记录难免会出错,进而给板材的管理带来不便。而伴随着现代图像识别技术的发展,以及图像识别在物流等领域的应用,给板材喷码识别提供了新的借鉴思路。而在当前的图像识别中,对图像进行特征提取,然后采用神经网络、SVM等算法对图像进行识别,是常用的思路。而本文则结合以往研究的思路,尝试将识别方法应用到板材的喷涂信息识别中,特别是数字字符和英文字符识别,以此为图像识别在板材管理的运用提供新思路。……

登录APP查看全文