基于高光谱成像技术的马铃薯外部缺陷检测

为了实现马铃薯的准确快速分级,提出基于高光谱成像技术的马铃薯外部缺陷检测方法。通过反射高光谱成像技术采集马铃薯干腐、表面碰伤、机械损伤、绿皮、孔洞以及发芽等6类外部缺陷样本及合格样本的高光谱图像。提取合格及各类缺陷样本感兴趣区域的光谱曲线并进行光谱特性分析,采用主成分分析法确定了5个特征波段（480、676、750、800和960nm）,以5个波段的主成分分析的第二主成分图像作为分类图像,识别率仅为61.52%;为了提高识别率,提出波段比算法与均匀二次差分算法相结合的方法,使缺陷识别率提高到95.65%。试验结果表明：通过高光谱成像技术可以准确有效地对常见马铃薯外部缺陷进行检测,为马铃薯在线无...

Full description

Saved in:
Bibliographic Details
Published in农业工程学报 Vol. 28; no. 21; pp. 221 - 228
Main Author 周竹 李小昱 陶海龙 高海龙
Format Journal Article
LanguageChinese
Published 华中农业大学工学院,武汉430070%浙江农林大学信息工程学院,临安 311300 2012
Subjects
主成分分析
图像处理
均匀二次差分
无损检测
波段比
马铃薯
高光谱图像
主成分分析
均匀二次差分
高光谱图像
图像处理
波段比
无损检测
马铃薯
Online AccessGet full text
ISSN1002-6819
DOI10.3969/j.issn.1002-6819.2012.21.031

Cover

More Information
Summary:为了实现马铃薯的准确快速分级,提出基于高光谱成像技术的马铃薯外部缺陷检测方法。通过反射高光谱成像技术采集马铃薯干腐、表面碰伤、机械损伤、绿皮、孔洞以及发芽等6类外部缺陷样本及合格样本的高光谱图像。提取合格及各类缺陷样本感兴趣区域的光谱曲线并进行光谱特性分析,采用主成分分析法确定了5个特征波段（480、676、750、800和960nm）,以5个波段的主成分分析的第二主成分图像作为分类图像,识别率仅为61.52%;为了提高识别率,提出波段比算法与均匀二次差分算法相结合的方法,使缺陷识别率提高到95.65%。试验结果表明：通过高光谱成像技术可以准确有效地对常见马铃薯外部缺陷进行检测,为马铃薯在线无损检测分级提供了参考。
Bibliography:principal component analysis; nondestructive testing; image processing; hyperspectral imaging; band ratio; symmetrical second difference algorithm; potato
11-2047/S
Zhou Zhu , Li Xiaoyu , Tao Hailong , Gao Hailong （1. College of Engineering, Huazhong Agricultural University, Wuhan 430070, China; 2. School of Information Engineering, Zhejiang Agriculture and Forestry University, Lin’an 311300, China）
In order to realize accurate and fast classification of potato, a novel detection method for potato external defects was proposed based on hyperspectral imaging technology. Potatoes with dry rot, normal and other six kinds of common defects were studied. First, region of interests spectral features of various defected areas were analyzed and principal component analysis method （PCA） was used to determined five characteristic bands （480、676、750、800 and 960 nm）. Next, PCA was performed again based on characteristic bands and the second principal component was used to classify defects of potatoes, the overall classifica
ISSN:1002-6819
DOI:10.3969/j.issn.1002-6819.2012.21.031