摘要
以"红富士"苹果为研究对象,提出基于高光谱成像技术结合图像分割技术的苹果表面缺陷的无损检测方法。采用高光谱图像采集系统(400 nm~1 000 nm)采集完好无损和表面有缺陷苹果的高光谱图像;对采集到的高光谱图像进行最小噪声分离变换,提取感兴趣区域的平均光谱反射率;采用图像分割技术提出苹果表面缺陷的无损检测方法。结果表明:采用最小噪声分离变换可有效地消除苹果高光谱图像中的噪声;在700 nm~800 nm以及900 nm~1 000 nm波段范围内完好无损和表面有缺陷的苹果的光谱反射率值具有明显的差异,同时选取特征波长717.98 nm处的光谱反射率值小于0.6以及982.59 nm处的光谱反射率值大于0.52作为区分苹果正常区域和表面缺陷区域的阈值条件,进一步利用阈值分割方法对80个完好无损苹果和40个表面有缺陷苹果的正确识别率分别为97.5%和95%。表明高光谱成像技术结合图像分割技术可实现苹果表面缺陷的无损检测。
The "Fuji"apples were taken as research object,the nondestructive detection of defect on apples was proposed based on hyperspectral imaging technology combined with image segmentation technology. And the hyperspectral imaging system was used to collect the hyperspectral image of apples with no defect and surface defect. After the minimum noise fraction(MNF)transform,the average spectral reflectance of the region of interest(ROI)was acquired. The nondestructive detection of defect on apples was proposed based on image segmentation technology and then applied on 80 no defect apples and 40 surface defect apples. Results showed that the noise of the hyperspectral image of apples can be effectively removed by MNF transform. And the no defect apples and surface defect apples had obvious reflectance value between 700 nm-800 nm and 900 nm-1 000 nm. The spectral reflectance at 717.98 nm was less than 0.6 and the spectral reflectance at 982.59 nm was greater than 0.52,which were both selected as the threshold condition to distinguish the normal region and the surface defect region of apples. The correct identification rates for no defect apples and surface defect apples reached to 97.5 % and 95 %,respectively. This study indicated that hyperspectral imaging technology combined with image segmentation technology was effective for identifying defect on apples.
引文
[1]刘民法,张令标,王松磊,等.近红外高光谱技术鉴别长枣表面的农药种类[J].食品研究与开发,2014,35(15):81-86
[2]杨晓玉,丁佳兴,房盟盟,等.基于可见/近红外高光谱成像技术的鸡蛋新鲜度无损检测[J].食品与机械,2017,33(11):131-136
[3]YAND D,HE D D,LU A X,et al.Detection of the Freshness State of Cooked Beef During Storage Using Hyperspectral Imaging[J].Applied Spectroscopy,2017,71(10):2286-2301
[4]SINEENART S,SONTISUK T.Non-destructive quality assessment of hens′eggs using hyperspectral images[J].Journal of Food Engineering,2017,215:97-103
[5]ERKINBAEV C,HENDERSON K,PALIWAL J.Discrimination of Gluten-Free Oats from Contaminants Using Near Infrared Hyperspectral Imaging Technique[J].Food Control,2017,80:197-203
[6]MO C,KIM G,KIM M S,et al.Fluorescence hyperspectral imaging technique for foreign substance detection on fresh-cut lettuce[J].Journal of the Science of Food&Agriculture,2017,97(12):3985-3993
[7]SIEDLISKA A,BARANOWSKI P,ZUBIK M,et al.Detection of pits in fresh and frozen cherries using a hyperspectral system in transmittance mode[J].Journal of Food Engineering,2017,215:61-71
[8]徐爽,何建国,马瑜,等.高光谱图像技术在水果品质检测中的研究进展[J].食品研究与开发,2013,43(10):4-8
[9]詹白勺,倪君辉,李军.高光谱技术结合CARS算法的库尔勒香梨可溶性固形物定量测定[J].光谱学与光谱分析,2014,34(10):2752-2757
[10]董金磊,郭文川.采后猕猴桃可溶性固形物含量的高光谱无损检测[J].食品科学,2015,36(16):101-106
[11]Li B C,HOU B L,ZHANG D W,et al.Pears characteristics(soluble solids content and firmness prediction,varieties)testing methods based on visible-near infrared hyperspectral imaging[J].Optik,2016,127:2624-2630
[12]刘燕德,吴明明,李轶凡,等.苹果可溶性固形物和糖酸比可见/近红外漫反射与漫透射在线检测对比研究[J].光谱学与光谱分析,2017,37(8):2424-2429
[13]管晓梅,杜军,张立人,等.基于高光谱技术的果糖检测优化算法和可视化方法[J].光电子·激光,2018,29(2):173-180
[14]赵杰文,刘剑华,陈全胜,等.利用高光谱图像技术检测水果轻微损伤[J].农业机械学报,2008,39(1):106-109
[15]张保华,黄文倩,李江波,等.基于高光谱成像技术和MNF检测苹果的轻微损伤[J].光谱学与光谱分析,2014,34(5):1367-1372
[16]田有文,程怡,王小奇,等.基于高光谱成像的苹果虫伤缺陷与果梗/花萼识别方法[J].农业工程学报,2015,31(4):325-331