基于高光谱图像技术羊肉表面污染无损检测  

Non-destructive detection of fecal contaminants on lamb meat using hyperspectral imaging technique

作　　者：思振华[1] 何建国[1] 刘贵珊[1] 王松磊[1] 贺晓光[1] 施健[2] 罗瑞明[1] 

机构地区：[1]宁夏大学农学院,宁夏银川750021 [2]宁夏大学物理电气信息学院,宁夏银川750021

出　　处：《食品与机械》2013年第5期75-79,共5页Food and Machinery

基　　金：国家科技支撑计划课题(编号:2012BAK17B07);国家自然科学基金(编号:31101306)

摘　　要：研究采用高光谱成像技术(光谱范围400～1 000nm)对羊肉表面肠溶物污染进行无损检测,在可见-近红外光谱范围内,对采集到的高光谱图像数据进行主成分分析,选取5种(430,540,574,622,660nm)大肠肠溶物及5种(420,473,540,574,660nm)小肠肠溶物特征波长,在特征波长下,对图像数据进行主成分分析,应用中值滤波、平方根变换、二值化和数字形态学对不同光谱范围内肠溶物进行特征提取,试验结果表明,基于全光谱范围对羊肉表面肠溶物的检测率为98.5%,基于特征波段光谱范围内对羊肉表面肠溶物的检测率为98.1%,证明利用高光谱图像技术对羊肉表面污染无损检测是可行的。Hyperspectral imaging technique operated in the visible and near infrared（Vis-NIR） region （400-1 000） was developed for noncontact measurement of fecal contaminants of lamb meat. Principal Component Analysis （PCA） was performed on hyperspectral imaging data in the Vis and Vis NIR region. Five wavelengths （430, 540, 574, 622 and 660 nm） were selected as the optimum wavelengths for fecal of intestinum crassum, and five wavelengths （420, 473, 540, 574 and 660 nm） were identified for fecal of intestinum tenue. In the key wavelengths, the data was analyzed by PCA. Fecal contaminants on lamb＇ s surface were segmented by median filtering, square root transformation, binarization and morphological imaging processing in the different region of spectral. Experimental results showed that the fecal on the lambs can be detected with an accuracy of 98.5 % in the region of Vis-NIR, and 98.1% in the optimum wavelengths. Therefore,hyperspecctral imaging technique can be used for the non-de-structive detection of fecal contaminants on lamb meat.

关 键 词：高光谱图像 羊肉 肠溶物 无损检测 

分 类 号：TS251.7[轻工技术与工程—农产品加工及贮藏工程] TP274[轻工技术与工程—食品科学与工程]