张学兰,李 军,孟范孔.一种基于机器视觉的纸病识别方法[J].中国造纸学报,2013,28(1):48-52 |
二维码(扫一下试试看!) |
一种基于机器视觉的纸病识别方法 |
A Paper Defects Detection Method Based on Machine Vision |
|
DOI:10.11981/j.issn.1000-6842.2013.01.48 |
中文关键词: 纸病图像 图像灰度 特征值 BP神经网络 |
Key Words:paper defect image image gray characteristic value BP neural network |
基金项目:国家重点基础研究发展规划(973计划,2010CB732205);国家科技支撑计划项目(2007BAF25B00)。 |
|
摘要点击次数: 4533 |
全文下载次数: 0 |
中文摘要: |
提出了一种基于图像灰度变换和BP神经网络的纸病识别算法。该算法是利用动态双阈值法和图像合成法对图像进行预处理,提取出纸病的特征值,再利用BP神经网络对纸病进行分类。结果表明,BP神经网络分类器可以很好地识别出纸病图像中的孔洞、脏点和褶皱,平均识别率达93.8%。 |
Abstract: |
A new paper defects recognition algorithm based on image gray transformation and BP neural network was put forward.The paper image was preprocessed, then the paper defects characteristic value was extracted, finally, the BP neural network was used to classify paper defects. Experimental results showed that this algorithm could successfully recognize a paper image that contains holes, spots and folds. The precision of the performance of the system reached 93.8%. |
查看全文 查看/发表评论 下载PDF阅读器 HTML |