Intelligent Identification and Classification of Common Bean Seed Varieties Based on Machine Vision

doi:10.11924/j.issn.1000-6850.casb2024-0295

Abstract

Abstract:

The aim of this study was to design a low-cost, efficient and non-destructive method for identifying and classifying common bean seeds based on machine vision (MV). In this study, image information of 2751 seeds of six varieties of common beans was collected, and based on image processing such as binarization, color extraction and morphological operations, nine features including color features, texture features and geometric features were extracted as the basis of classification, and K-nearest neighbor (KNN), random forest (RF), and support vector machine (SVM) classification models were established to classify the varieties of bean seeds. After comparing the confusion matrix, accuracy and F₁ value of the three classification models, it was found that the SVM model outperformed the other two classification models, with a classification accuracy and F₁ value of 97.7% and 0.977, respectively. The results of the study show that accurate identification and classification of common bean seeds can be achieved using MV.

Key words: common bean, classification, machine vision, image processing, intelligent recognition, machine learning, support vector machine, feature extraction

LI Shujia, SUN Laijun, MENG Yahao, WANG Yichen, LI Xiaoxu, FENG Guojun, YANG Fengyan. Intelligent Identification and Classification of Common Bean Seed Varieties Based on Machine Vision[J]. Chinese Agricultural Science Bulletin, 2025, 41(4): 156-164.

Figures/Tables 12

References 33

[1]	张东昱, 宋学林, 王进. 4个菜豆品种种子特征特性及活力比较[J]. 耕作与栽培, 2020, 40(4):22-27.
[2]	SHOUCHE S, RASTOGI R, BHAGWAT S, et al. Shape analysis of grains of Indian wheat varieties[J]. Computers electronics in agriculture, 2001, 33(1):55-76.
[3]	邓小琴, 朱启兵, 黄敏. 融合光谱、纹理及形态特征的水稻种子品种高光谱图像单粒鉴别[J]. 激光与光电子学进展, 2015, 52(2):128-134.
[4]	BAEK I, KUSUMANINGRUM D, KANDPAL L M, et al. Rapid measurement of soybean seed viability using Kernel- Based multispectral image analysis[J]. Sensors, 2019, 19(2):271.
[5]	DANA W, IVO W. Computer image analysis of seed shape and seed color for flax cultivar description[J]. Computers and electronics in agriculture, 2008, 61(2):126-135.
[6]	成芳, 应义斌. 机器视觉技术在农作物种子质量检验中的应用研究进展[J]. 农业工程学报, 2001(6):175-179.
[7]	TEMPELAERE A, KETELAERE B D, HE J, et al. An introduction to artificial intelligence in machine vision for postharvest detection of disorders in horticultural products[J]. Postharvest biology and technology, 2023, 206:112576.
[8]	LV X, ZHANG X, GAO H, et al. When crops meet machine vision: A review and development framework for a low-cost nondestructive online monitoring technology in agricultural production[J]. Agriculture communications, 2024, 2(1):100029.
[9]	PANDEY N, KRISHNA S, SHARMA S. Automatic seed classification by shape and color features using machine vision technology[J]. International journal of computer applications technology research, 2013, 2(2):208-213.
[10]	祝保林. 基于机器视觉的小桐子种子识别分类研究[D]. 郑州: 河南大学, 2020.
[11]	黄琼, 杨红云, 万颖. 基于特征数据的水稻种子分类识别方法[J]. 江苏农业学报, 2021, 37(1):8-15.
[12]	KOKLU M, OZKAN I A. Multiclass classification of dry beans using computer vision and machine learning techniques[J]. Computers and electronics in agriculture, 2020, 174:105507.
[13]	ROPELEWSKA E, RUTKOWSKI K P. Cultivar discrimination of stored apple seeds based on geometric features determined using image analysis[J]. Journal of stored products research, 2021, 92:101804.
[14]	DE VITIS M, HAY F R, DICKIE J B, et al. Seed storage: Maintaining seed viability and vigor for restoration use[J]. Restoration ecology, 2020, 28:249-255.
[15]	郝建平, 杨锦忠, 杜天庆. 基于图像处理的玉米品种的种子形态分析及其分类研究[J]. 中国农业科学, 2008(4):994-1002.
[16]	ALI A, QADRI S, MASHWANI W K, et al. Machine learning approach for the classification of corn seed using hybrid features[J]. International journal of food properties, 2020, 23(1):1110-1124.
[17]	ALIIEV E. Automatic phenotyping test of sunflower seeds[J]. Helia, 2020, 43(72):51-66.
[18]	XU M, MIAO D, YU G, et al. A bearing dynamic model based on novel Gaussian-filter waviness characterizing method for vibration response analysis[J]. Tribology international, 2024, 194:109433.
[19]	LU Y, DUANMU L, ZHAI Z, et al. Application and improvement of Canny edge-detection algorithm for exterior wall hollowing detection using infrared thermal images[J]. Energy and buildings, 2022, 274:112421.
[20]	SU R, SUN C, ZHANG C, et al. A new method for linear feature and junction enhancement in 2D images based on morphological operation, oriented anisotropic Gaussian function and Hessian information[J]. Pattern recognition, 2014, 47(10):3193-3208.
[21]	JI Y, SUN L, LI Y, et al. Non-destructive classification of defective potatoes based on hyperspectral imaging and support vector machine[J]. Infrared physics & technology, 2019, 99:71-79.
[22]	MAO S, LI Y, MA Y, et al. Automatic cucumber recognition algorithm for harvesting robots in the natural environment using deep learning and multi-feature fusion[J]. Computers and electronics in agriculture, 2020, 170:105254.
[23]	卢蓉, 范勇, 陈念年, 王俊波. 一种提取目标图像最小外接矩形的快速算法[J]. 计算机工程, 2010, 36(21):178-180. doi: 10.3969/j.issn.1000-3428.2010.21.064
[24]	GUPTA A, CHAKROBORTY S, GHOSH S, et al. A machine learning model for multi-class classification of quenched and partitioned steel microstructure type by the k-nearest neighbor algorithm[J]. Computational materials science, 2023, 228:112321.
[25]	LIAW A, WIENER M. Classification and regression by randomForest[J]. R news, 2002, 2(3):18-22.
[26]	HEARST M A, DUMAIS S T, OSUNA E, et al. Support vector machines[J]. Intelligent systems their applications, 1998, 13(4):18-28.
[27]	HAYATI R, MUNAWAR A, LUKITANINGSIH E, et al. Combination of PCA with LDA and SVM classifiers: A model for determining the geographical origin of coconut in the coastal plantation, Aceh Province, Indonesia[J]. Case studies in chemical and environmental engineering, 2024, 9:100552.
[28]	ROOIJ M, WEEDA W. Cross-validation: A method every psychologist should know[J]. Advances in methods practices in psychological science, 2020, 3(2):248-263.
[29]	VISA S, RAMSAY B, RALESCU A L, et al. Confusion matrix-based feature selection[J]. Midwest artificial intelligence and cognitive science conference, 2011, 710:120-127.
[30]	ALAPATI Y K, SINDHU K. Combining clustering with classification: A technique to improve classification accuracy[J]. Lung cancer, 2016, 32(57):3.
[31]	DEVRIES Z, LOCKE E, HODA M M, et al. Using a national surgical database to predict complications following posterior lumbar surgery and comparing the area under the curve and F₁-score for the assessment of prognostic capability[J]. The spine journal, 2021, 21(7):1135-1142.
[32]	FUGARA A, AHMADLOU M, SHABEEB A R R, et al. Spatial mapping of groundwater springs potentiality using grid search-based and genetic algorithm-based support vector regression[J]. Geocarto international, 2020, 37(1):284-303.
[33]	THARWAT A. Parameter investigation of support vector machine classifier with kernel functions[J]. Knowledge information systems, 2019, 61(3):1269-1302.

品种			本体颜色						花纹颜色						纵横比		种子大小/cm²		花纹数目
品种			R		G		B		R		G		B		纵横比		种子大小/cm²		花纹数目
丰冠	1		161.79		127.19		101.14		95.41		123.97		139.16		1.4		1.16		25
	2		159.57		131.27		106.48		105.92		127.84		126.52		1.49		1.18		20
	3		167.86		139.37		111.16		105.13		134.03		140.28		1.45		1.15		27
	4		175.36		143.89		115.29		109.46		134.72		132.86		1.41		0.79		24
	平均值		166.15		135.43		108.52		103.98		130.14		134.71		1.44		1.07		24
青冠	1	191.31		155.31		107.9		104		147.71		150.78		1.46		0.82		1
	2	162.77		131.38		87.96		84.41		123.84		149.43		1.46		0.7		2
	3	137.07		115.62		84.76		80.15		113.79		134.31		1.4		0.8		1
	4	153.19		127		89		82.83		120.64		143.91		1.49		0.74		2
	平均值	161.09		132.33		92.41		87.85		126.50		144.61		1.45		0.77		1.5
霞冠	1	120.85		109.02		92.96		87.08		102.76		114.22		1.65		1.18		1
	2	118.42		115.75		102.98		94.2		104.4		106.32		1.76		1.01		2
	3	141.11		120.19		93.15		92.38		121.88		137.85		1.59		1.27		2
	4	110.98		95.56		81.95		77.51		93.46		110.94		1.49		1.17		2
	平均值	122.84		110.13		92.76		87.79		105.63		117.33		1.62		1.16		1.75
白冠	1	201.14		186.93		151.5		147.62		163.82		165.17		1.27		0.8		0
	2	187.69		184.42		156.88		150.17		163.13		162.18		1.33		0.76		0
	3	191.36		180.08		148.76		144.58		160.23		162.14		1.29		0.74		0
	4	208.5		198.15		160.9		150.1		164.43		164.97		1.22		0.58		0
	平均值	197.17		187.40		154.51		148.12		162.90		163.62		1.28		0.72		0
玉冠	1	177.72		146.72		109.96		102.26		131.99		150.83		1.72		0.72		1
	2	142.57		121.09		95.7		89.61		104.36		127.61		1.6		0.52		2
	3	169.85		148.81		117.26		108.93		129.2		146.6		1.89		0.72		1
	4	151.85		138.96		114.44		93.59		107.2		125.91		1.71		0.56		1
	平均值	160.50		138.90		109.34		98.60		118.19		137.74		1.73		0.63		1.25
盛冠	1	113.76		95.16		83.49		73.19		83.65		104.23		1.47		0.91		16
	2	130.55		110.21		91.14		79.89		99.18		118.96		1.35		0.81		20
	3	124.74		103.84		85.42		76.47		94.06		115.51		1.38		0.86		26
	4	135.94		112.65		94.68		81.19		97.73		121.87		1.32		0.76		23
	平均值	126.25		105.47		88.68		77.69		93.66		115.14		1.38		0.84		21.25

品种			本体颜色						花纹颜色						纵横比		种子大小/cm²		花纹数目
品种			R		G		B		R		G		B		纵横比		种子大小/cm²		花纹数目
丰冠	1		161.79		127.19		101.14		95.41		123.97		139.16		1.4		1.16		25
	2		159.57		131.27		106.48		105.92		127.84		126.52		1.49		1.18		20
	3		167.86		139.37		111.16		105.13		134.03		140.28		1.45		1.15		27
	4		175.36		143.89		115.29		109.46		134.72		132.86		1.41		0.79		24
	平均值		166.15		135.43		108.52		103.98		130.14		134.71		1.44		1.07		24
青冠	1	191.31		155.31		107.9		104		147.71		150.78		1.46		0.82		1
	2	162.77		131.38		87.96		84.41		123.84		149.43		1.46		0.7		2
	3	137.07		115.62		84.76		80.15		113.79		134.31		1.4		0.8		1
	4	153.19		127		89		82.83		120.64		143.91		1.49		0.74		2
	平均值	161.09		132.33		92.41		87.85		126.50		144.61		1.45		0.77		1.5
霞冠	1	120.85		109.02		92.96		87.08		102.76		114.22		1.65		1.18		1
	2	118.42		115.75		102.98		94.2		104.4		106.32		1.76		1.01		2
	3	141.11		120.19		93.15		92.38		121.88		137.85		1.59		1.27		2
	4	110.98		95.56		81.95		77.51		93.46		110.94		1.49		1.17		2
	平均值	122.84		110.13		92.76		87.79		105.63		117.33		1.62		1.16		1.75
白冠	1	201.14		186.93		151.5		147.62		163.82		165.17		1.27		0.8		0
	2	187.69		184.42		156.88		150.17		163.13		162.18		1.33		0.76		0
	3	191.36		180.08		148.76		144.58		160.23		162.14		1.29		0.74		0
	4	208.5		198.15		160.9		150.1		164.43		164.97		1.22		0.58		0
	平均值	197.17		187.40		154.51		148.12		162.90		163.62		1.28		0.72		0
玉冠	1	177.72		146.72		109.96		102.26		131.99		150.83		1.72		0.72		1
	2	142.57		121.09		95.7		89.61		104.36		127.61		1.6		0.52		2
	3	169.85		148.81		117.26		108.93		129.2		146.6		1.89		0.72		1
	4	151.85		138.96		114.44		93.59		107.2		125.91		1.71		0.56		1
	平均值	160.50		138.90		109.34		98.60		118.19		137.74		1.73		0.63		1.25
盛冠	1	113.76		95.16		83.49		73.19		83.65		104.23		1.47		0.91		16
	2	130.55		110.21		91.14		79.89		99.18		118.96		1.35		0.81		20
	3	124.74		103.84		85.42		76.47		94.06		115.51		1.38		0.86		26
	4	135.94		112.65		94.68		81.19		97.73		121.87		1.32		0.76		23
	平均值	126.25		105.47		88.68		77.69		93.66		115.14		1.38		0.84		21.25

处理	项目	本体颜色			花纹颜色			纵横比	种子大小/cm²	花纹数目
处理	项目	R	G	B	R	G	B	纵横比	种子大小/cm²	花纹数目
丰冠	最小值	127	98	84	73	89	99	1.23	1.16	15
	最大值	202	163	133	130	144	156	1.74	1.36	29
	均值	163	131	105	101	123	131	1.47	1.02	21
	方差	13	10	6	7	9	9	0.07	0.12	4.34
青冠	最小值	101	86	68	60	75	94	1.16	0.45	0
	最大值	191	155	111	105	147	157	1.79	1.12	2
	均值	137	115	85	78	107	129	1.47	0.72	1
	方差	11	9	6	6	9	10	0.08	0.11	0.03
霞冠	最小值	89	83	72	62	74	81	1.34	0.67	0
	最大值	161	138	109	104	133	149	1.82	1.34	3
	均值	123	108	88	81	102	117	1.59	1.02	2.3
	方差	13	10	11	10	10	12	0.08	0.10	0.05
白冠	最小值	154	142	102	95	128	130	1.05	0.36	0
	最大值	214	204	171	159	168	166	1.58	0.89	0
	均值	190	179	144	138	159	161	1.31	0.61	0
	方差	8	10	12	12	5	3	0.07	0.09	0
玉冠	最小值	122	97	76	67	89	110	1.33	0.35	1
	最大值	186	164	128	118	140	157	2.01	0.91	2
	均值	151	130	102	93	116	137	1.71	0.66	1
	方差	11	10	8	7	8	8	0.09	0.09	0.49
盛冠	最小值	99	83	72	66	74	90	1.05	0.49	16
	最大值	157	126	103	96	117	143	1.51	1.07	27
	均值	126	103	87	79	94	117	1.33	0.74	21
	方差	10	7	4	5	8	9	0.06	0.10	4.48

处理	项目	本体颜色			花纹颜色			纵横比	种子大小/cm²	花纹数目
处理	项目	R	G	B	R	G	B	纵横比	种子大小/cm²	花纹数目
丰冠	最小值	127	98	84	73	89	99	1.23	1.16	15
	最大值	202	163	133	130	144	156	1.74	1.36	29
	均值	163	131	105	101	123	131	1.47	1.02	21
	方差	13	10	6	7	9	9	0.07	0.12	4.34
青冠	最小值	101	86	68	60	75	94	1.16	0.45	0
	最大值	191	155	111	105	147	157	1.79	1.12	2
	均值	137	115	85	78	107	129	1.47	0.72	1
	方差	11	9	6	6	9	10	0.08	0.11	0.03
霞冠	最小值	89	83	72	62	74	81	1.34	0.67	0
	最大值	161	138	109	104	133	149	1.82	1.34	3
	均值	123	108	88	81	102	117	1.59	1.02	2.3
	方差	13	10	11	10	10	12	0.08	0.10	0.05
白冠	最小值	154	142	102	95	128	130	1.05	0.36	0
	最大值	214	204	171	159	168	166	1.58	0.89	0
	均值	190	179	144	138	159	161	1.31	0.61	0
	方差	8	10	12	12	5	3	0.07	0.09	0
玉冠	最小值	122	97	76	67	89	110	1.33	0.35	1
	最大值	186	164	128	118	140	157	2.01	0.91	2
	均值	151	130	102	93	116	137	1.71	0.66	1
	方差	11	10	8	7	8	8	0.09	0.09	0.49
盛冠	最小值	99	83	72	66	74	90	1.05	0.49	16
	最大值	157	126	103	96	117	143	1.51	1.07	27
	均值	126	103	87	79	94	117	1.33	0.74	21
	方差	10	7	4	5	8	9	0.06	0.10	4.48

标签	品种	训练集	测试集	总计
1	丰冠	216	86	302
2	青冠	435	185	620
3	霞冠	202	101	303
4	白冠	502	201	703
5	玉冠	308	133	441
6	盛冠	262	120	382
总计		1925	826	2751