The New Lens on Seed Testing: AI, ML and Image Analysis
834
Keywords:
Digital seed testing, artificial intelligence, image analysis, seed quality, hyperspectral imaging, machine learningAbstract
Seed testing is undergoing a paradigm shift with the integration of advanced digital technologies
that augment conventional methodologies by offering greater speed, accuracy, and scalability. This review examines
the transformative impact of innovations such as automation, artificial intelligence (AI), and advanced image
analysis on contemporary seed quality assessment. AI-based approaches, particularly machine learning and
deep learning models, are increasingly being employed to automate complex and labour-intensive tasks—including
seed sorting, germination prediction, and defect detection—that were traditionally reliant on expert visual evaluation.
Image analysis techniques, notably hyperspectral and multispectral imaging, enable rapid and non-destructive
assessment of both external and internal seed attributes, thereby substantially improving diagnostic precision. In
parallel, the adoption of edge computing, Internet of Things (IoT) sensors, and cloud-based platforms has enabled
real-time and decentralized seed testing, extending analytical capabilities from centralized laboratories to fieldlevel applications. These digital frameworks enhance data integration, traceability, and decision-making efficiency
across seed production systems, breeding programmes, and quality control laboratories. Moreover, high-throughput
digital phenotyping tools are accelerating plant breeding by generating detailed, high-resolution trait data, thereby
reducing selection cycles and improving the efficiency of crop improvement programmes. Despite these
advancements, the widespread adoption of digital seed testing technologies necessitates the development of
standardized protocols, robust validation frameworks, and appropriate regulatory oversight to ensure data reliability,
interoperability, and broader applicability. Collectively, these emerging innovations are redefining seed testing
practices and positioning the discipline as a critical enabler of sustainable, precision-oriented, and data-driven
agriculture.
Downloads
References
1. 1. KUMAR A, S SINGH AND N MEHTA (2022). Applications of
2. artificial intelligence and image analysis in seed quality testing.
3. Journal of Agricultural Informatics, 13(1): 10-20.
4. 2. COLMER J, CM O NEILL, R WELLS, A BOSTROM, D
5. REYNOLDS, D WEBSALE (2020). SeedGerm: a cost-effective
6. phenotyping platform for automated seed imaging and
7. machine-learning based phenotypic analysis of crop seed
8. germination. New Phytologist, 228(2): 778-793.
9. 3. FENG L, S ZHU, F LIU, Y HE, Y BAO AND C ZHANG (2019).
10. Hyperspectral imaging for seed quality and safety inspection:
11. a review. Plant Methods, 15: 91.
12. 4. BHARGAVA A, A SACHDEVA, K SHARMA, MH ALSHARIF, P
13. UTHANSAKUL AND M UTHANSAKUL (2024). Hyperspectral
14. imaging and its applications: a review. Heliyon, 10(12): e33208.
15. 5. GUO D, Q ZHU, M HUANG, Y GUO AND J QIN (2017). Model
16. updating for the classification of different varieties of maize
17. seeds from different years by hyperspectral imaging coupled
18. with a pre-labeling method. Computers and Electronics in
19. Agriculture, 142: 1-8.
20. 6. SUN J, S JIANG, H MAO, X WU AND Q LI (2015).
21. Classification of black beans using visible and near infrared
22. hyperspectral imaging. International Journal of Food
23. Properties, 19(8): 1687-1695.
24. 7. YU Z, H FANG, Q ZHANGJIN, C MI, X FENG AND Y HE (2021).
25. Hyperspectral imaging technology combined with deep
26. learning for hybrid okra seed identification. Biosystems
27. Engineering, 212: 46-61.
28. 8. AN D, L ZHANG, Z LIU, J LIU AND Y WEI (2023). Advances in
29. infrared spectroscopy and hyperspectral imaging combined
30. with artificial intelligence for the detection of cereals quality.
31. Critical Reviews in Food Science and Nutrition, 63(29): 9766-
32. 9796.
33. 9. ZHAO X, H QUE, X SUN, Q ZHU AND M HUANG (2022).
34. Hybrid convolutional network based on hyperspectral imaging
35. for wheat seed varieties classification. Infrared Physics and
36. Technology, 125: 104270.
37. 10. LI B, J SUN, Y LIU, L SHI, Y ZHONG AND P WU (2024).
38. Multi-level data fusion strategy based on spectral and image
39. information for identifying varieties of soybean seeds.
40. Spectrochimica Acta Part A: Molecular and Biomolecular
41. Spectroscopy, 322: 124815.
42. 11. SUGANTHI M AND JGR SATHIASEELAN (2022). A novel
43. feature extraction method for identifying quality seed selection.
44. International Journal of Intelligent Engineering, 8(1): 1-10.
45. 12. GHIMIRE A, SH KIM, A CHO, N JANG, S AHN, MS ISLAM
46. (2023). Automatic evaluation of soybean seed traits using RGB
47. image data and a Python algorithm. Plants, 12(17): 3078.
48. 13. JIN B, C ZHANG, L JIA, Q TANG, L GAO, G ZHAO AND H QI
49. (2022). Identification of rice seed varieties based on nearinfrared hyperspectral imaging technology combined with deep
50. learning. ACS Omega, 7(6): 4735-4749.
51. 14. AZNAN A, C GONZALEZ VIEJO, A PANG AND S FUENTES
52. (2021). Computer vision and machine learning analysis of
53. commercial rice grains: a potential digital approach for
54. consumer perception studies. Sensors, 21(19): 6354.
55. 15. JEYARAJ PR, SP ASOKAN AND ERS NADAR (2022).
56. Computer-assisted real-time rice variety learning using deep
57. learning network. Rice Science, 29(5): 489-498.
58. 16. WHAN AP, AB SMITH, CR CAVANAGH, JP RAL, LM SHAW,
59. CA HOWITT AND L BISCHOF (2014). GrainScan: a low-cost,
60. fast method for grain size and colour measurements. Plant
61. Methods, 10(1): 23.
62. 17. WIJAYANTO AK, A JUNAEDI, AA SUJASWARA, MBR
63. KHAMID, LB PRASETYO, C HONGO AND H KUZE (2023).
64. Machine learning for precise rice variety classification in
65. tropical environments using UAV-based multispectral sensing.
66. Agri. Engineering, 5(4): 2000-2019.
67. 18. LI Z, X JIANG, X JIA, X DUAN, Y WANG AND J MU (2022).
68. Classification method of significant rice pests based on deep
69. learning. Agronomy, 12(9): 2096.
70. 19. HUANG M, C HE, Q ZHU AND J QIN (2016). Maize seed
71. variety classification using the integration of spectral and image
72. features combined with feature transformation based on
73. hyperspectral imaging. Applied Sciences, 6(6): 183.
74. 20. ZHAO Y, S ZHU, C ZHANG, X FENG, L FENG AND Y HE
75. (2018). Application of hyperspectral imaging and
76. chemometrics for variety classification of maize seeds. RSC
77. Advances, 8: 1337-1345.
78. 21. WANG QJ, SY ZHANG, SF DONG, GC ZHANG, J YANG, R LI
79. AND HQ WANG (2020). Pest24: a large-scale very small
80. object data set of agricultural pests for multi-target detection.
81. Computers and Electronics in Agriculture, 175: 105585.
82. 22. KUMAR R, A GUPTA, S SRIVASTAVA, G DEVI, VK SINGH,
83. SK GOSWAMI (2020). Diagnosis and detection of seed-borne
84. fungal phytopathogens. In: KUMAR R AND A GUPTA (Eds.),
85. Seed-Borne Diseases of Agricultural Crops: Detection,
86. Diagnosis & Management. Singapore: Springer Nature Pte
87. Ltd, 107-142.
88. 23. BALA A (2020). Non-parasitic seed disorders of major
89. agricultural crops. In: KUMAR R AND A GUPTA (Eds.), SeedBorne Diseases of Agricultural Crops: Detection, Diagnosis &
90. Management. Singapore: Springer Nature Pte Ltd, 809-820.
91. 24. MARTINELLI F, R SCALENGHE, S DAVINO, S PANNO, G
92. SCUDERI, P RUISI (2015). Advanced methods of plant
93. disease detection: a review. Agronomy for Sustainable
94. Development, 35: 1-25.
95. 25. WHETTON RL, KL HASSALL, TW WAINE AND AM
96. MOUAZEN (2018). Hyperspectral measurements of yellow
97. rust and fusarium head blight in cereal crops: Part 1:
98. Laboratory study. Biosystems Engineering, 166: 101-115.
99. 26. GOWEN AA, CP O’DONNELL, PJ CULLEN, G DOWNEY AND
100. JM FRIAS (2007). Hyperspectral imaging-an emerging
101. process analytical tool for food quality and safety control.
102. Trends in Food Science & Technology, 18(12): 590-598.
103. 27. MAHESH S, DS JAYAS, J PALIWAL AND NDG WHITE (2015).
104. Hyperspectral imaging to classify and monitor quality of
105. agricultural materials. Journal of Stored Products Research,
106. 61: 17-26.
107. 28. YAN K, MKC SHISHER AND Y SUN (2023). A transfer
108. learning-based deep convolutional neural network for detection
109. of Fusarium wilt in banana crops. Agri. Engineering, 5(4): 2381-
110. 2394.
111. 29. YIN J, LI W, J SHEN, C ZHOU, S LI AND J SUO (2025). A
112. diffusion-based detection model for accurate soybean disease
113. identification in smart agricultural environments. Plants, 14(5):
114. 675.
115. 30. XU P, L FU, K XU, WB SUN, Q TAN AND Y ZHANG (2023).
116. Investigation into maize seed disease identification based on
117. deep learning and multi-source spectral information fusion
118. techniques. Journal of Food Composition and Analysis, 119:
119. 105254.
120. 31. QIU R, C YANG, A MOGHIMI, M ZHANG, BJ STEFFENSON
121. AND CD HIRSCH (2019). Detection of Fusarium head blight
122. in wheat using a deep neural network and color imaging.
123. Remote Sensing, 11(22): 2658.
124. 32. SASTRY DVSR, HD UPADHYAYA AND CLL GOWD (2014).
125. Determination of physical properties of chickpea seeds and
126. their relevance in germplasm collections. Indian Journal of
127. Plant Genetic Resources, 27(1): 1-9.
128. 33. KADIR MFA, NAN YUSRI, M RIZON, ARB MAMAT, AA JAMAL
129. AND M MAKHTAR (2015). Automatic mango detection using
130. texture analysis and randomised Hough transform. Applied
131. Mathematical Sciences, 9: 6427-6436.
132. 34. GONZALEZ R AND R WOODS (2017). Digital Image
133. Processing, 4th ed. Pearson International, pp. 464-540.
134. 35. SAFADI T, M KANG, ICC LEITE AND B VIDAKOVIC (2016).
135. Wavelet-based spectral descriptors for detection of damage
136. in sunflower seeds. International Journal of Wavelets,
137. Multiresolution and Information Processing, 14: 1650027.
138. 36. GIRI A, V SAGAN, H ALIFU, A MAIWULANJIANG, S SARKAR,
139. B ROY AND FB FRITSCHI (2024). A wavelet decomposition
140. method for estimating soybean seed composition with
141. hyperspectral data. Remote Sensing, 16(23): 4594.
142. 37. SUREKHA R, R SHOBARANI AND GVS GEORGE (2019).
143. Seed classification using multi-feature extraction. International
144. Journal of Innovative Technology and Exploring Engineering,
145. 8(8S): 635-639.
146. 38. CINAR I AND M KOKLU (2019). Classification of rice varieties
147. using artificial intelligence methods. International Journal of
148. Intelligent Systems and Applications in Engineering.
149. 39. AGARWAL D, SWETA AND B PATTIMA (2023). Machine
150. learning approach for the classification of wheat grains. Smart
151. Agricultural Technology, 3: 100136.
152. 40. SAHA D AND A MANICKAVASAGAN (2022). Chickpea varietal
153. classification using deep convolutional neural networks with
154. transfer learning. Journal of Food Process Engineering, 45(3):
155. e13975.
156. 41. YANG S, L ZHENG, P HE (2021). High-throughput soybean
157. seeds phenotyping with convolutional neural networks and
158. transfer learning. Plant Methods, 17: 50.
159. 42. QIU G, E LU, H LU, S XU, F ZENG AND Q SHUI (2018).
160. Single-kernel FT-NIR spectroscopy for detecting supersweet
161. corn (Zea mays L. saccharata Sturt) seed viability with
162. multivariate data analysis. Sensors, 18(4): 1010.
163. 43. LEE HS, YA JEON, YY LEE, GA LEE, S RAVEENDAR AND
164. KH MA (2017). Large-scale screening of intact tomato seeds
165. for viability using near infrared reflectance spectroscopy
166. (NIRS). Sustainability, 9: 618.
167. 44. KANDPAL LM, S LOHUMI, MS KIM, JS KANG AND BK CHO
168. (2016). Near-infrared hyperspectral imaging system coupled
169. with multivariate methods to predict viability and vigor in
170. muskmelon seeds. Sensors and Actuators B: Chemical, 229:
171. 534-544.
172. 45. JOOSEN RVL, J KODDE, LAJ WILLEMS, W LIGTERINK,
173. LHW VAN DER PLAS AND HWM HILHORST (2010).
174. GERMINATOR: a software package for high-throughput
175. scoring and curve fitting of Arabidopsis seed germination. Plant
176. Journal, 62(1): 148-159.
177. 46. HE X, X FENG, D SUN, F LIU, Y BAO AND Y HE (2019).
178. Rapid and nondestructive measurement of rice seed vitality
179. of different years using near-infrared hyperspectral imaging.
180. Molecules, 24(12): 2227.
181. 47. OLESEN MH, P NIKNESHAN, A SHRESTHA AND B BOELT
182. (2011a). Viability and germination of castor seeds (Ricinus
183. communis L.) determined by multispectral imaging. Seed
184. Science and Technology, 39(2): 389-400.
185. 48. WANG Y AND S SONG (2024). Detection of sweet corn seed
186. viability based on hyperspectral imaging combined with firefly
187. algorithm optimized deep learning. Frontiers in Plant Science,
188. 15: Article 1361309.
189. 49. WAKHOLI C, LM KANDPAL, H LEE, H BAE, E PARK, MS
190. KIM, C MO, WH LEE AND BK CHO (2018). Rapid assessment
191. of corn seed viability using short wave infrared line-scan
192. hyperspectral imaging and chemometrics. Sensors and
193. Actuators B: Chemical, 255: 498-507.
194. 50. SINGH K, H DUDDU, S VAIL, I PARKIN AND S SHIRTLIFFE
195. (2021). UAV-based hyperspectral imaging technique to
196. estimate canola (Brassica napus L.) seedpods maturity.
197. Canadian Journal of Remote Sensing, 47(1): 1-20.
198. 51. FENG Y, X ZHAO, R TIAN, C LIANG, J LIU AND X FAN (2024).
199. Research on an intelligent seed-sorting method and sorter
200. based on machine vision and lightweight YOLOv5n. Agronomy,
201. 14(9): 1953.
202. 52. SINGH SK, R JHA, S PANDEY, C MOHAN, CHETNA, S
203. GHOSH, SK SINGH, S KUMARI AND A SINGH (2025).
204. Artificial intelligence-based tools for next-generation seed
205. quality analysis. Crop Design, 4(1): 100094.
206. 53. LI C, H LI, L LIAO, Z LIU AND Y DONG (2023). Real-time
207. seed sorting system via 2D information entropy-based CNN
208. pruning and TensorRt acceleration. IET Image Processing,
209. 17(6): 1694-1708.
210. 54. KOPPAD D, KV SUMA AND N NAGARAJAPPA (2024).
211. Automated seed classification using state-of-the-art
212. techniques. SN Computer Science, 5(5): 511.
213. 55. GHAMARI S, AM BORGHEI, H RABBANI, J KHAZAEI AND
214. F BASATI (2010). Modeling the terminal velocity of agricultural
215. seeds with artificial neural networks. African Journal of
216. Agricultural Research, 5(5): 389-398.
217. 56. JIAN F (2022). A general model to predict germination and
218. safe storage time of crop seeds. Journal of Stored Products
219. Research, 99: 102041.
220. 57. SANKARAN S, A MISHRA, R EHSANI AND C DAVIS (2016).
221. Separation of viable and non-viable tomato (Solanum
222. lycopersicum L.) seeds using single seed near-infrared
223. spectroscopy. Computers and Electronics in Agriculture, 127:
224. 633-640.
225. 58. ZHANG Y, Y WANG AND X LI (2024). Nondestructive detection
226. of sunflower seed vigor and moisture content based on
227. hyperspectral imaging and chemometrics. Foods, 13(9): 1320.
228. 59. MAHLEIN AK (2016). Plant disease detection by imaging
229. sensors - Parallels and specific demands for precision
230. agriculture and plant phenotyping. Plant Disease, 100(2): 241-
231. 251.
232. 60. SINGH P, A NAYYAR, S SINGH AND A KAUR (2020).
233. Classification of wheat seeds using image processing and
234. fuzzy clustered random forest. International Journal of
235. Agricultural Resources, Governance and Ecology, 16(2): 123.
236. 61. FUENTES A, S YOON, SC KIM AND DS PARK (2017). A
237. robust deep-learning-based detector for real-time tomato plant
238. diseases and pests recognition. Sensors, 17(9): 2022.
239. 62. KNIBBE WJ, L AFMAN, S BOERSMA, MJ BOGAARDT, J
240. EVERS, F VAN EVERT AND A DE WIT (2022). Digital twins
241. in the green life sciences. NJAS Impact in Agricultural and
242. Life Sciences, 94(1): 249-279.
243. 63. CARBONELL IM (2016). The ethics of big data in big
244. agriculture. Internet Policy Review, 5(1).
245. 64. KAMILARIS A AND FX PRENAFETA-BOLDÚ (2018). Deep
246. learning in agriculture: A survey. Computers and Electronics
247. in Agriculture, 147(4): 70-90.
248. 65. STAHL BC (2021). Artificial intelligence for a better future: An
249. ecosystem perspective on the ethics of AI and emerging digital
250. technologies. Springer Nature, p. 124.
251. 66. KLERKX L, E JAKKU AND P LABARTHE (2019). A review of
252. social science on digital agriculture, smart farming and
253. agriculture 4.0: New contributions and a future research
254. agenda. NJAS - Wageningen Journal of Life Sciences, 90-
255. 91: 100315.
256. 67. SHIMELES A, A VERDIER-CHOUCHANE AND A BOLY
257. (2018). Introduction: Understanding the challenges of the
258. agricultural sector in Sub-Saharan Africa. In: Building a
259. Resilient and Sustainable Agriculture in Sub-Saharan Africa.
260. Palgrave Macmillan, Cham.
261. 68. POWELL AA (2006). Seed vigour and its assessment. In:
262. Handbook of Seed Science and Technology. Food Products
263. Press, pp. 603-648.
264. 69. AOSA (2019). Rules for Testing Seeds. Association of Official
265. Seed Analysts, USA.
266. 70. U.S DEPARTMENT OF AGRICULTURE - AGRICULTURAL
267. MARKETING SERVICE (2023). Seed Regulatory and Testing
268. Division.
269. 71. FOOD AND AGRICULTURE ORGANIZATION OF THE
270. UNITED NATIONS (2021). Guidelines for the establishment
271. and management of seed testing laboratories. Joint FAO and
272. ISTA Handbook.
273. 72. SRINIVASAIAH R, M MEENAKSHI, R CHENNEGOWDA AND
274. S JANKATTI (2023). Analysis and prediction of seed quality
275. using machine learning. International Journal of Electrical and
276. Computer Engineering, 13: 5770.
277. 73. RIJSWIJK K, L KLERKX, M BACCO, F BARTOLINI, E
278. BULTEN, L DEBRUYNE, J DESSEIN, I SCOTTI AND G
279. BRUNORI (2021). Digital transformation of agriculture and
280. rural areas: A socio-cyber-physical system framework to
281. support responsibilisation. Journal of Rural Studies, 85: 79-
282. 90.
283. 74. INTERNATIONAL SEED TESTING ASSOCIATION (ISTA)
284. (2023). International Rules for Seed Testing: 2023 Edition.
285. ISTA.
286. 75. DIBBERN T, LAS ROMANI AND SMFS MASSRUHÁ (2024).
287. Main drivers and barriers to the adoption of digital agriculture
288. technologies. Smart Agricultural Technology, 8: 100459.