基于NIR高光谱图像的冬枣损伤早期检测研究

食品科学 ›› 0, Vol. ›› Issue (): 0-0.

• 安全检测 • 下一篇

基于NIR高光谱图像的冬枣损伤早期检测研究

孙世鹏¹,彭俊¹,李瑞¹,朱兆龙¹,Vázquez-Arellano Manuel²,傅隆生¹

1. 西北农林科技大学
2. University of Hohenheim

收稿日期:2016-06-21 修回日期:2016-11-16 出版日期:2017-01-25 发布日期:2017-01-16
通讯作者: 傅隆生 E-mail:fulsh@nwafu.edu.cn
基金资助:
枣果的振动采收机理研究;果实的自动分级研究

Early detection of mechanical damage in Chinese winter jujube (Zizyphus jujuba Mill. cv. Dongzao) using NIR hyperspectral images

¹, ¹,li rui ¹,Vázquez-Arellano Manuel¹,Long-Sheng FU¹

Received:2016-06-21 Revised:2016-11-16 Online:2017-01-25 Published:2017-01-16
Contact: Long-Sheng FU E-mail:fulsh@nwafu.edu.cn

摘要/Abstract

摘要： 冬枣在机械振动采摘过程中掉落以及分选过程中碰撞，容易损伤，但是初期不可见，无法采用普通彩色相机检测。为了对冬枣损伤进行早期检测，本文采用NIR高光谱图像技术对损伤区域成像。为了决定哪种掉落高度下，损伤检测更容易，根据掉落高度划分冬枣为3种损伤区域。针对高光谱成像波段多的特点，分别采用连续投影法（Successive Projections Algorithm，SPA）、相关特征选择算法（Correlation based Feature Selection，CFS）、一致性（Consistency）算法选择冬枣损伤的特征波段，对提取的特征波段分别应用K-邻近（k-Nearest Neighbor，k-NN）、朴素贝叶斯（Naive Bayes，NB）、支持向量机（Support Vector Machine，SVM）3种分类方法进行损伤区域识别。试验结果表明所有方法选择的一致特征波段在1353和1691 nm附近。Consistency选择的特征波段在SVM分类器下损伤检测效果最好，达到95.16%，一致特征波长在NB分类器下分类识别正确率达到84.26%，验证了一致波长的有效性，为多光谱成像技术实现在线检测冬枣损伤奠定基础。

关键词: 高光谱成像, 冬枣, 轻微损伤, 检测, 特征波段

Abstract: Winter jujube (Zizyphus jujuba Mill.) is an important fruit in China due to its good taste and abundant nutrition. Its fruits are sensitive and can easily develop brown spots after suffering mechanical stress during mechanical harvesting and postharvest handling. The damage cannot be detected easily by machine vision in very early stages of maturity. Thus an NIR hyperspectral imaging system was used to detect mechanical damage. The damage was produced by dropping fruit from different height in order to estimate the moment in which the damage could be effectively detected. For reducing the dimensionality of hyperspectral data, three feature selection methods (Successive Projections Algorithm, SPA; Correlation Based Feature Selection, CFS; Consistency) were used. In addition, three classifiers (k-Nearest Neighbor, k-NN; Naive Bayes, NB; Support Vector Machine, SVM) were evaluated to segment the pixels of the jujubes into two classes: damaged and non-damaged. Results revealed there are two common wavebands 1353 nm (No. 130) and 1691 nm (No. 232) in all the feature selection methods. Besides, SVM offering the best performance which reached 95.16% with the selected features set by the Consistency. NB offering similarly performance which reached 84.26% with the two common wavebands. Hence, this work lays the foundations for on-line detecting early damage caused by mechanical stress.

Key words: Chinese jujube, Hyperspectral imaging, Feature selction, Slight damage, Detection

中图分类号:

TP391.4

孙世鹏彭俊李瑞朱兆龙 Vázquez-Arellano Manuel 傅隆生. 基于NIR高光谱图像的冬枣损伤早期检测研究[J]. 食品科学, 0, (): 0-0.

li rui Vázquez-Arellano Manuel Long-Sheng FU. Early detection of mechanical damage in Chinese winter jujube (Zizyphus jujuba Mill. cv. Dongzao) using NIR hyperspectral images[J]. FOOD SCIENCE, 0, (): 0-0.

参考文献

[1] GAO Q H, WU P T, LIU J R, et al. Physico-chemical properties and antioxidant capacity of different jujube (Ziziphus jujuba Mill.) cultivars grown in loess plateau of China [J]. Scientia Horticulturae, 2011, 130(1): 67-72. DOI: 10.1016/j.scienta.2011.06.005.
[2] 中华人民共和国质量监督检验总局, 中国国家标准化委员会. GB/T 22345—2008 鲜枣质量等级 [S].北京: 中国标准出版社, 2009.
[3] 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, (05): 1367-1372.
[4] 魏新华, 吴姝, 范晓冬, 等. 基于高光谱成像分析的冬枣微观损伤识别[J]. 农业机械学报, 2015, 46(3): 242-246. DOI: 10.6041/j.issn.1000-1298.2015.03.035.
[5] WANG J, NAKANO K, OHASHI S. Nondestructive evaluation of jujube quality by visible and near-infrared spectroscopy[J]. LWT - Food Science and Technology, 2011, 44(4): 119-1125. DOI: 10.1016/j.lwt.2010.11.012.
[6] YU K, ZHAO Y, LI X, et al. Identification of crack features in fresh jujube using Vis/NIR hyperspectral imaging combined with image processing[J]. Computers and Electronics in Agriculture, 2014, 103(1-10). DOI: 10.1016/j.compag.2014.01.016.
[7] 辛世华, 何建国, 王松磊, 等. 红枣表面损伤的特征光谱提取[J]. 食品科学, 2013, 34(8): 145-148.
[8] 胡鹏程, 孙晔, 吴海伦, 等. 高光谱图像对白萝卜糠心的无损检测[J]. 食品科学, 2015, 36(12): 171-176.
[9] FU L, OKAMOTO H, SHIBATA Y, et al. Distinguishing overripe berries of Japanese blue honeysuckle using hyperspectral imaging analysis[J]. Engineering in Agriculture, Environment, and Food, 2014, 7(1): 22-27. DOI: 10.1016/j.eaef.2013.12.004
[10] 薛建新,张淑娟, 张晶晶. 壶瓶枣自然损伤的高光谱成像检测 [J]. 农业机械学报, 2015, 46(7): 220-226. DOI: 10.6041/j.issn.1000-1298.2015.07.032.
[11] VELEZ RIVERA N, GOMEZ-SANCHIS J, CHANONA-P REZ J, et al. Early detection of mechanical damage in mango using NIR hyperspectral images and machine learning [J]. Biosystems Engineering, 2014, 122:91-98. DOI: 10.1016/j.biosystemseng.2014.03.009.
[12] RAJKUMAR P, WANG N, ELMASRY G, et al. Studies on banana fruit quality and maturity stages using hyperspectral imaging[J]. Journal Of Food Engineering, 2012, 108(1): 194-200.DOI: 10.1016/j.jfoodeng.2011.05.002.
[13] 张初, 刘飞, 孔汶汶, 等. 利用近红外高光谱图像技术快速鉴别西瓜种子品种[J]. 农业工程学报, 2013,( 20): 270-277.
[14] SOARES S F C, GOMES A A, ARAUJO M C U, et al. The successive projections algorithm[J]. TrAC Trends in Analytical Chemistry, 2013, 42:84-98. DOI: 10.1016/j.trac.2012.09.006.
[15] HALL M A. Correlation based Feature Selection for Machine Learning[D]. Hamilton. University of California, 2000: 15-55.
[16] LIU H, SETIONO R. A Probabilistic Approach to Feature Selection - A Filter Solution[C]// proceedings of the International Conference on Machine Learning, 2010.
[17] GALVAO R K H, ARAUJO M C U, SILVA E C, et al. Cross-validation for the selection of spectral variables using the successive projections algorithm[J]. Journal of the Brazilian Chemical Society, 2007, 18(8): 1580-1584.
[18] 闭小梅, 闭瑞华. KNN算法综述[J]. 科技创新导报, 2009, (14): 31.
[19] BECKER B, KOHAVI R, DAN S. Visualizing the Simple Bayesian Classifier[J]. Integration of Data Mining and Data Visualization, 1998, 1-10.
[20] CORTES C, VAPNIK V. Support-Vector Networks[J]. Machine Learning, 1995, 20: 273-297.
[21] FRIEND J, RHODES M J C. Recent advances in the biochemistry of fruits and vegetables[M]. Academic Press, 1981.
[22] LU R, CEN H, HUANG M, et al. Spectral absorption and scattering properties of normal and bruised apple tissue[J]. Transactions of the ASABE, 2010, 53(1): 263-269.
[23] 蔡天净, 唐瀚. Savitzky-Golay平滑滤波器的最小二乘拟合原理综述[J]. 数字通信, 2011, (01): 63-68.
[24] YANG C, LEE W S, WILLIAMSON J G. Classification of blueberry fruit and leaves based on spectral signatures[J]. Biosystems Engineering, 2012, 113(4): 351-362. DOI: 10.1016/j.biosystemseng.2012.09.009.

基于NIR高光谱图像的冬枣损伤早期检测研究

Early detection of mechanical damage in Chinese winter jujube (Zizyphus jujuba Mill. cv. Dongzao) using NIR hyperspectral images

RichHTML

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	徐祥，谷贵章，尚佳宇，张进杰，林邦楚，谢红丰，徐大伦. 低盐雪菜中呈味核苷酸关联物的检测及贮藏过程中的变化[J]. 食品科学, 2024, 45(4): 315-322.
[2]	徐悦，陈海艺，周梦含，刘艺璇，郭红莲. 纳米Cu/多糖复合抗菌膜的制备与表征及其对冬枣黑斑病的防治效果[J]. 食品科学, 2024, 45(3): 125-133.
[3]	王艳，包士雷，孙立瑞，窦博鑫，辛嘉英，张娜. 电化学技术在食品检测中的应用研究进展[J]. 食品科学, 2024, 45(3): 295-306.
[4]	衡新蕊，王天亮，姚云平，李昌模. 冷榨亚麻籽油特征风味成分分析[J]. 食品科学, 2024, 45(2): 211-217.
[5]	马立鑫，吴玮，许骞，尹丽梅，韩恩，白竣文，蔡健荣. 表面增强拉曼光谱法同时检测果汁中的啶虫脒和福美双[J]. 食品科学, 2024, 45(2): 283-289.
[6]	占可，陈季旺，徐言，倪杨帆，刘言，邹圣碧. 基于近红外光谱特征的冷冻小龙虾鲜度快速检测方法[J]. 食品科学, 2024, 45(2): 299-307.
[7]	祁兴普，董骐玮，朱麟菲，邹婷婷，郑义，张婧怡，李倩. 基于金纳米花的双信号适配体传感器检测红酒中真菌毒素[J]. 食品科学, 2024, 45(2): 308-314.
[8]	潘良文，宁雪，王强，蔡一村，许镇坚，张晨. 火鸡源性成分实时聚合酶链式反应检测方法ISO标准研制[J]. 食品科学, 2024, 45(1): 217-224.
[9]	宁亚维，周泓鑫，杨正，马俊美，刘茁，张岩，李强. UPLC-MS/MS法检测3 种食品中松仁过敏原[J]. 食品科学, 2024, 45(1): 247-253.
[10]	韩子馨，张丽丽，张博，邹方磊，尚楠. 新型无损检测技术在番茄品质检测中的研究与应用进展[J]. 食品科学, 2024, 45(1): 289-300.
[11]	李桢桢，尹明雨，王红丽，王锡昌. 水产品肌肉组织微观结构变化及其检测方法研究进展[J]. 食品科学, 2023, 44(9): 278-286.
[12]	王元清，李莎，周巧，覃池，李建龙，李琴，胡凯弟，刘书亮. 粮油原料中农药使用与残留现状及快速检测技术研究进展[J]. 食品科学, 2023, 44(9): 287-296.
[13]	王帅，杨艳歌，吴占文，李红娜，李涛，孙冬梅，袁飞. 重组酶聚合酶扩增、重组酶介导等温扩增及酶促重组等温扩增技术在食源性致病菌快速检测中的研究进展[J]. 食品科学, 2023, 44(9): 297-305.
[14]	彭悦，文春露，杨春杰，蒋圣启，丁武. 利用青海弧菌Q67评估5 种农药的毒性作用[J]. 食品科学, 2023, 44(9): 15-24.
[15]	何晓婷, 陈子键, 黄松, 肖泽苗, 刘佳, 钟敏, 王弘, 沈玉栋, 徐振林. 基于纳米抗体的胶体金免疫层析法快速检测蔬菜中腐霉利[J]. 食品科学, 2023, 44(8): 307-316.