基于纹理图像的点张计数算法设计与分析

doi:10.3969/j.issn.1000-1158.2017.04.07

摘要
图/表
参考文献(12)
相关文章 (15)

全文: PDF (974 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要针对手工点张耗时长、精度低的问题,提出了一种应用图像纹理分析的纸张计数算法。用高精度图像摄取装置对纸张图像进行采集,对原始纹理图像采用高斯滤波法去除噪点,大大提高了计算机处理速度。引进方向滤波与Gabor小波变换结合的方法对纹理图像进行增强提升了点张准确度;针对图像中难以处理的纸张堆叠问题,提出了计数修正算法。对不同厚度及张数纸堆进行了点数试验。对于0.3mm的堆叠纸张,点张准确率100%; 对于0.2mm的堆叠纸张,点张准确率达99%以上。试验结果表明该算法具有较好的计数能力,计数准确高。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	梁海毅
	王纪森
	黎雯
	孙灵远

关键词 ：计量学, 纸张计数, 纹理图像, 高斯滤波, Gabor变换

Abstract：A paper counting algorithm based on the paper texture analysis is presented. The aim of the optimal design is to replace the manual paper counting with time-consuming and low precision, which collect the paper image with high precision image pickup device. To improve the speed of processing speed, Gauss filtering method was used to remove noise. Then directional filtering method and Gabor wavelet transform are combined to enhance the texture image. A counting correction algorithm is proposed according to the paper stack case, for the 0.3 mm stack of paper, the algorithm points to an accuracy rate is 100%, for the 0.2 mm stack of paper, the point of the accuracy rate is more than 99%. Test results show that the algorithm has good counting ability and high count accuracy.

Key words： metrology paper counting texture image Gauss filter Gabor wavelet transform

收稿日期: 2015-05-12 发布日期: 2017-06-16

PACS:

TB96

通讯作者: 梁海毅 E-mail: nicam116@aliyun.com

作者简介: 梁海毅（1979-）,男,广东顺德人,西北工业大学在读博士研究生,主要研究方向为自动化检测与控制、电液控制系统及飞机地面控制系统。nicam116@aliyun.com.

引用本文:

梁海毅,王纪森,黎雯,孙灵远. 基于纹理图像的点张计数算法设计与分析[J]. 计量学报, 2017, 38(4): 416-419.
LIANG Hai-yi,WANG Ji-sen,LI Wen,SUN Ling-yuan. Design and Reseach of Paper Counting Algorithem Based on Texture Image. Acta Metrologica Sinica, 2017, 38(4): 416-419.

链接本文:

http://jlxb.china-csm.org:81/Jwk_jlxb/CN/10.3969/j.issn.1000-1158.2017.04.07 或 http://jlxb.china-csm.org:81/Jwk_jlxb/CN/Y2017/V38/I4/416