Non-destructive identification of artificially aged alfalfa seeds using multispectral imaging analysis

doi:10.11686/cyxb2021198

Abstract

Abstract:

Aging during the storage of seeds reduces seed vitality and causes serious economic losses to the agricultural industry， and has become one of the biggest factors involved in decreased seed vigor. Distinction between aged and viable seeds is of high importance in alfalfa seed planting and production， but the existing methods are time-consuming or destructive. Therefore， a rapid and non-destructive screening method to distinguish aged and viable seeds is not only very necessary in seed testing and the alfalfa seed industry， but also potentially useful in alfalfa seed research. In this study， we collected data of both morphological features and spectral traits of alfalfa seeds using multispectral imaging （MSI） technology. Then， we evaluated three multivariate analysis methods： linear discriminant analysis （LDA）， support vector machines （SVM） and normalized canonical discriminant analysis （nCDA）， to classify seeds artificially aged for 0， 3， 6 and 14 days， and predict viable seeds which could germinate. It was found that the mean light reflectance at 470-660 nm differed significantly between non-aged and aged seeds. The LDA model based on a “hold-out method” provided accuracies of 93.0%-97.7% in distinguishing aged seeds from non?aged seeds， and 75.3%-91.7% in distinguishing the different groups of aged seeds. Corresponding values for the SVM model were a little lower， being 92.4%-94.9% and 68.7%-78.8%， respectively. The nCDA model also exhibited achieved aged seed discrimination with an accuracy of 88.0%-98.0%. Finally， viable seeds could be distinguished from dead seeds in all the categories of aged seeds， with accuracies of 98.7% and 92.1% in LDA and SVM analysis， respectively， while the accuracy of nCDA in predicting the germination of aged seeds ranged from 90% to 99%. This study showed that MSI could successfully distinguish aged seeds， and also predict germination of seeds. In summary， we demonstrated a nondestructive， rapid and high-throughput approach to screen both aged and viable seeds in alfalfa， and showed that MSI together with multivariate analysis is promising as a new tool for application in seed testing and field planting of alfalfa seeds.

Key words: aged seeds, multispectral imaging, multivariate analysis, alfalfa, non-destructive identification

Xue-meng WANG, Xin HE, Han ZHANG, Rui SONG, Pei-sheng MAO, Shan-gang JIA. Non-destructive identification of artificially aged alfalfa seeds using multispectral imaging analysis[J]. Acta Prataculturae Sinica, 2022, 31(7): 197-208.

Figures/Tables 11

References 44

1	Yang W M， Liu G， Ou Q H， et al. Infrared spectroscopy of naturally aged legume seeds. Laser and Optoelectronics Progress， 2018， 55（12）： 513-522.
	杨卫梅，刘刚，欧全宏，等. 豆类种子自然老化过程的红外光谱研究. 激光与光电子学进展， 2018， 55（12）： 513-522.
2	Xia F， Cheng H， Chen L， et al. Influence of exogenous ascorbic acid and glutathione priming on mitochondrial structural and functional systems to alleviate aging damage in oat seeds. BMC Plant Biology， 2020， 20（1）： 104.
3	Neelesh K， Arvind A， Siddiqui M A， et al. Physiological and biochemical changes during seed deterioration in aged seeds of rice （Oryza sativa L.）. American Journal of Plant Physiology， 2011， 6（1）： 28-35.
4	McDonald M B. Seed deterioration： Physiology， repair and assessment. Seed Science and Technology， 1999， 27（1）： 177-237.
5	Liu J， Liu G， Li S J， et al. Study on artificial aged rice seed based on infrared spectroscopy. Seed， 2021， 40（1）： 17-22.
	刘杰，刘刚，李姝洁，等. 人工老化水稻种子的红外光谱研究. 种子， 2021， 40（1）： 17-22.
6	Jin T， Liu W， Liu C H. Research on rapid nondestructive detection of moisture content of beef jerky based on multispectral imaging technology. Journal of Anhui Agricultural Sciences， 2021， 49（2）： 204-205.
	金涛，刘伟，刘长虹. 基于多光谱成像技术的牛肉干水分含量快速无损检测研究. 安徽农业科学， 2021， 49（2）： 204-205.
7	ElMasry G， Mandour N， Rejaie S， et al. Recent applications of multispectral imaging in seed phenotyping and quality monitoring-an overview. Sensors， 2019， 19（5）： 1-32.
8	Xu X， Ma H， Wang Y， et al. Rapid and nondestructive identification of wheat varieties with multispectral imaging technology. Chinese Agricultural Science Bulletin， 2019， 35（15）： 14-19.
	许学，马卉，王钰，等. 基于多光谱成像技术的小麦品种快速无损鉴定. 中国农学通报， 2019， 35（15）： 14-19.
9	Hu X， Yang L， Zhang Z， et al. Differentiation of alfalfa and sweet clover seeds via multispectral imaging. Seed Science and Technology， 2020， 48（1）： 83-99.
10	Olesen M H， Nikneshan P， Shrestha S， et al. Viability prediction of Ricinus cummunis L. seeds using multispectral imaging. Sensors， 2015， 15（2）： 4592-4604.
11	Baek I， Kusumaningrum D， Kandpal L M， et al. Rapid measurement of soybean seed viability using kernel-based multispectral image analysis. Sensors， 2019， 19（2）： 271.
12	Rego C H Q， Franca S F， Gomes J F G， et al. Using multispectral imaging for detecting seed-borne fungi in cowpea. Agriculture， 2020， 10（8）： 361.
13	Vresak M， Olesen M H， Gislum R， et al. The use of image-spectroscopy technology as a diagnostic method for seed health testing and variety identification. PLoS One， 2016， 11（3）： 1-10.
14	National Animal Husbandry General Station. Regional specialization of forage seed production. Beijing： China Agriculture Press， 2018.
	全国畜牧总站. 牧草种子专业化生产的地域性. 北京：中国农业出版社， 2018.
15	Shi F L， Shi F L， Gao C P， et al. Vigor change of artificial aging alfalfa seeds and ISSR markers. Chinese Journal of Grassland， 2015， 37（6）： 30-34.
	石凤玲，石凤翎，高翠萍，等. 人工老化紫花苜蓿种子活力变化与ISSR标记. 中国草地学报， 2015， 37（6）： 30-34.
16	Qu Y Y， Huang X， Xu Q H， et al. Effects of artificial aging on physiological and biochemical characteristics of Brassica alboglabra bailey seeds. Seed， 2020， 39（8）： 24-29.
	屈煜莹，黄鑫，徐强辉，等. 人工老化处理对芥蓝种子生理生化特性的影响. 种子， 2020， 39（8）： 24-29.
17	Ji J H， Meng C M， He Q F. Influence of different aging times on physiological and biochemical characteristics of cotton. Seed， 2017， 36（11）： 14-17.
	姬俊华，孟超敏，何庆飞. 不同老化处理时间对棉花种子生理生化的影响. 种子， 2017， 36（11）： 14-17.
18	Hu X， Yang L， Zhang Z. Non-destructive identification of single hard seed via multispectral imaging analysis in six legume species. Plant Methods， 2020， 16（1）： 116.
19	Yang Z S， Guo L， Luo X， et al. A PCA based band selection algorithm of hyperspectral image. Microelectronics and Computer， 2006， 12（21）： 72-74.
	杨诸胜，郭雷，罗欣，等. 一种基于主成分分析的高光谱图像波段选择算法. 微电子学与计算机， 2006， 12（21）： 72-74.
20	Hu R， Zhu X， Zhu Y， et al. Robust SVM with adaptive graph learning. World Wide Web， 2020， 23（3SI）： 1945-1968.
21	Pornpanomchai C， Jongsriwattanaporn S， Pattanakul T， et al. Image analysis on color and texture for chili （Capsicum frutescence） seed germination. Science， Engineering and Health Studies， 2020， 14（3）： 169-183.
22	Lurstwut B， Pornpanomchai C. Image analysis based on color， shape and texture for rice seed （Oryza sativa L.） germination evaluation. Agriculture and Natural Resources， 2017， 51（5）： 383-389.
23	Boelt B， Shrestha S， Salimi Z， et al. Multispectral imaging-a new tool in seed quality assessment. Seed Science Research， 2018， 28（3）： 222-228.
24	Li Z H， Chen Y， Ye D Y， et al. CIELAB colour space quantification-based evaluation of capsule development and seed vigour in Nicotiana tabacum L. Chinese Tobacco Science， 2015， 36（4）： 24-28.
	李振华，陈尧，叶定勇，等. 基于CIELAB颜色空间判别烤烟蒴果发育与种子活力变化. 中国烟草科学， 2015， 36（4）： 24-28.
25	Cong X X. Study on the determination of seed vigour of perennial ryegrass using multispectral imaging technology. Beijing： China Agricultural University， 2020.
	丛晓翔. 利用多光谱成像技术测定多年生黑麦草种子活力的研究. 北京：中国农业大学， 2020.
26	Yang L， Zhang Z， Hu X. Cultivar discrimination of single alfalfa （Medicago sativa L.） seed via multispectral imaging combined with multivariate analysis. Sensors， 2020， 20（22）： 6575.
27	Lei X， Liu W， Zhao J， et al. Comparative physiological and proteomic analysis reveals different involvement of proteins during artificial aging of Siberian wildrye seeds. Plants， 2020， 9（10）： 1370.
28	Shrestha S， Deleuran L C， Gislum R. Classification of different tomato seed cultivars by multispectral visible-near infrared spectroscopy and chemometrics. Journal of Spectral Imaging， 2016， a1（5）： 1-9.
29	ElMasry G， Mandour N， Wagner M， et al. Utilization of computer vision and multispectral imaging techniques for classification of cowpea （Vigna unguiculata） seeds. Plant Methods， 2019， 15（1）： 24.
30	Salimi Z， Boelt B. Classification of processing damage in sugar beet （Beta vulgaris） seeds by multispectral image analysis. Sensors， 2019， 19（10）： 2360.
31	Weng H， Tian Y， Wu N， et al. Development of a low-cost narrow band multispectral imaging system coupled with chemometric analysis for rapid detection of rice false smut in rice seed. Sensors， 2020， 20（4）： 1209.
32	Martins B V D J， Mascarin G M， Aparecida S S L C， et al. Multispectral and X-ray images for characterization of Jatropha curcas L. seed quality. Plant Methods， 2021， 17（1）： 1-24.
33	Franca S F， Queiroz R C H， Guilhien G J F， et al. Detection of Drechslera avenae （Eidam） sharif ［Helminthosporium avenae （Eidam）］ in black oat seeds （Avena strigosa Schreb） using multispectral imaging. Sensors， 2020， 20（12）： 3343-3353.
34	Liu C， Liu W， Lu X， et al. Non-destructive discrimination of conventional and glyphosate-resistant soybean seeds and their hybrid descendants using multispectral imaging and chemometric methods. The Journal of Agricultural Science， 2016， 154（1）： 1-12.
35	Galletti P A， Carvalho M E A， Hirai W Y， et al. Integrating optical imaging tools for rapid and non-invasive characterization of seed quality： Tomato （Solanum lycopersicum L.） and carrot （Daucus carota L.） as study cases. Frontiers in Plant Science， 2020， 11： 577851.
36	Witjes H， Rijpkema M， van der Graaf M， et al. Multispectral magnetic resonance image analysis using principal component and linear discriminant analysis. Journal of Agnetic Resonance Imaging， 2003， 17（2）： 261-269.
37	Mortensen A K， Gislum R， Jørgensen J R， et al. The use of multispectral imaging and single seed and bulk near-infrared spectroscopy to characterize seed covering structures： Methods and applications in seed testing and research. Agriculture， 2021， 11（4）： 301.
38	Wang X Z， Lu Q， Zhang X D， et al. Non-destructive detection of cucumber seeds vigor based on hyperspectral imaging. Jiangsu Journal of Agricultural Sciences， 2019， 35（5）： 1197-1202.
	王新忠，卢青，张晓东，等. 基于高光谱图像的黄瓜种子活力无损检测. 江苏农业学报， 2019， 35（5）： 1197-1202.
39	Rahman A， Cho B. Assessment of seed quality using non-destructive measurement techniques： A review. Seed Science Research， 2016， 26（4）： 285-305.
40	Wu J Z， Li H， Zhang H D， et al. Nondestructive determination of natural aging stage of wheat seeds using near infrared spectroscopy. Spectroscopy and Spectral Analysis， 2019， 39（3）： 751-755.
	吴静珠，李慧，张鹤冬，等. 小麦种子自然老化程度的近红外光谱无损识别. 光谱学与光谱分析， 2019， 39（3）： 751-755.
41	Chen L L， Xia F S， Mao P S， et al. Application of near infrared spectroscopy （NIR） technique in hard seed testing research. Infrared Technology， 2015， 37（9）： 793-798.
	陈玲玲，夏方山，毛培胜，等. 近红外光谱技术在种子硬实检测中的研究进展. 红外技术， 2015， 37（9）： 793-798.
42	Turki A， Baskin C C. Determination of seed viability of eight wild Saudi Arabian species by germination and X-ray tests. Saudi Journal of Biological Sciences， 2017， 24（4）： 822-829.
43	He X， Feng X， Sun D， et al. Rapid and nondestructive measurement of rice seed vitality of different years using near-infrared hyperspectral imaging. Molecules， 2019， 24（12）： 2227.
44	Kandpal L M， Lohumi S， Kim M S， et al. Near-infrared hyperspectral imaging system coupled with multivariate methods to predict viability and vigor in muskmelon seeds. Sensors and Actuators B： Chemical， 2016， 229： 534-544.

特征 Feature	CK	A3	A6	A14
面积 Area （mm²）	3.05±0.43	3.40±0.50**	3.54±0.56**	3.48±0.54**
长度 Length （mm）	2.49±0.24	2.64±0.25**	2.70±0.26**	2.68±0.29**
宽度 Width （mm）	1.62±0.12	1.70±0.14**	1.73±0.16**	1.71±0.14**
长宽比 Ratio of width and length	0.66±0.07	0.65±0.06	0.64±0.06	0.64±0.08
紧实度圆 Compactness circle	0.63±0.06	0.62±0.07	0.62±0.06	0.62±0.09
紧实度椭圆 Compactness ellipse	0.99±0.01	0.99±0.01**	0.99±0.01	0.99±0.01**
形状参数a Beta shape_a	1.51±0.12	1.47±0.15*	1.49±0.13	1.48±0.16
形状参数b Beta shape_b	1.44±0.12	1.40±0.13*	1.41±0.12	1.41±0.15
垂直偏度 Vertical skewness	-0.04±0.03	-0.04±0.03	-0.04±0.04	-0.04±0.03
颜色参数L* CIELab L*	48.50±4.99	46.65±4.25**	45.98±4.55**	45.46±4.90**
颜色参数A* CIELab A*	9.50±2.00	11.90±2.82**	12.79±2.87**	13.38±3.18**
颜色参数B* CIELab B*	30.39±3.44	29.14±2.83**	28.85±2.53**	29.21±2.82*
饱和度 Saturation	31.73±2.92	31.46±2.29	31.55±1.97	32.26±1.87
色调 Hue	1.26±0.09	1.18±0.10**	1.15±0.10**	1.14±0.11**

特征 Feature	CK	A3	A6	A14
面积 Area （mm²）	3.05±0.43	3.40±0.50**	3.54±0.56**	3.48±0.54**
长度 Length （mm）	2.49±0.24	2.64±0.25**	2.70±0.26**	2.68±0.29**
宽度 Width （mm）	1.62±0.12	1.70±0.14**	1.73±0.16**	1.71±0.14**
长宽比 Ratio of width and length	0.66±0.07	0.65±0.06	0.64±0.06	0.64±0.08
紧实度圆 Compactness circle	0.63±0.06	0.62±0.07	0.62±0.06	0.62±0.09
紧实度椭圆 Compactness ellipse	0.99±0.01	0.99±0.01**	0.99±0.01	0.99±0.01**
形状参数a Beta shape_a	1.51±0.12	1.47±0.15*	1.49±0.13	1.48±0.16
形状参数b Beta shape_b	1.44±0.12	1.40±0.13*	1.41±0.12	1.41±0.15
垂直偏度 Vertical skewness	-0.04±0.03	-0.04±0.03	-0.04±0.04	-0.04±0.03
颜色参数L* CIELab L*	48.50±4.99	46.65±4.25**	45.98±4.55**	45.46±4.90**
颜色参数A* CIELab A*	9.50±2.00	11.90±2.82**	12.79±2.87**	13.38±3.18**
颜色参数B* CIELab B*	30.39±3.44	29.14±2.83**	28.85±2.53**	29.21±2.82*
饱和度 Saturation	31.73±2.92	31.46±2.29	31.55±1.97	32.26±1.87
色调 Hue	1.26±0.09	1.18±0.10**	1.15±0.10**	1.14±0.11**

特征 Feature	A3	A6	A14
面积 Area （mm²）	3.40±0.50	3.54±0.56	3.48±0.54
长度 Length （mm）	2.64±0.25	2.70±0.26	2.68±0.29
宽度 Width （mm）	1.70±0.14	1.73±0.16	1.71±0.14
长宽比 Ratio of width and length	0.65±0.06	0.64±0.06	0.64±0.08
紧实度圆 Compactness circle	0.62±0.07	0.62±0.06	0.62±0.09
紧实度椭圆 Compactness ellipse	0.99±0.01	0.99±0.01	0.99±0.01
形状参数a Beta shape_a	1.47±0.15	1.49±0.13	1.48±0.16
形状参数b Beta shape_b	1.40±0.13	1.41±0.12	1.41±0.15
垂直偏度 Vertical skewness	-0.04±0.03	-0.04±0.04	-0.04±0.03
颜色参数L* CIELab L*	46.65±4.25	45.98±4.55	45.46±4.90*
颜色参数A* CIELab A*	11.90±2.82	12.79±2.87*	13.38±3.18**
颜色参数B* CIELab B*	29.14±2.83	28.85±2.53	29.21±2.82
饱和度 Saturation	31.46±2.29	31.55±1.97	32.26±1.87**
色调 Hue	1.18±0.10	1.15±0.10	1.14±0.11**

特征 Feature	A3	A6	A14
面积 Area （mm²）	3.40±0.50	3.54±0.56	3.48±0.54
长度 Length （mm）	2.64±0.25	2.70±0.26	2.68±0.29
宽度 Width （mm）	1.70±0.14	1.73±0.16	1.71±0.14
长宽比 Ratio of width and length	0.65±0.06	0.64±0.06	0.64±0.08
紧实度圆 Compactness circle	0.62±0.07	0.62±0.06	0.62±0.09
紧实度椭圆 Compactness ellipse	0.99±0.01	0.99±0.01	0.99±0.01
形状参数a Beta shape_a	1.47±0.15	1.49±0.13	1.48±0.16
形状参数b Beta shape_b	1.40±0.13	1.41±0.12	1.41±0.15
垂直偏度 Vertical skewness	-0.04±0.03	-0.04±0.04	-0.04±0.03
颜色参数L* CIELab L*	46.65±4.25	45.98±4.55	45.46±4.90*
颜色参数A* CIELab A*	11.90±2.82	12.79±2.87*	13.38±3.18**
颜色参数B* CIELab B*	29.14±2.83	28.85±2.53	29.21±2.82
饱和度 Saturation	31.46±2.29	31.55±1.97	32.26±1.87**
色调 Hue	1.18±0.10	1.15±0.10	1.14±0.11**

模型 Model	指标 Index	CK vs A3	CK vs A6	CK vs A14	A3 vs A6	A3 vs A14	A6 vs A14	G vs NG
SVM	准确度Accuracy	94.9	92.4	92.4	78.8	71.1	68.7	92.1
	敏感性Sensitivity	97.0	95.2	96.0	84.6	66.5	72.4	94.8
	特异性Specificity	92.8	89.7	88.7	73.2	75.6	64.6	89.5
LDA	准确度Accuracy	97.7	93.0	97.3	91.7	82.8	75.3	98.7
	敏感性Sensitivity	99.3	95.6	98.0	94.4	79.7	73.0	98.0
	特异性Specificity	96.1	90.5	96.8	88.3	86.0	78.2	99.3