Real-Time Deployment of MobileNetV3 Model in Edge Computing Devices Using RGB Color Images for Varietal Classification of Chickpea
Source abstract
Abstract
Chickpeas are one of the most widely consumed pulses globally because of their high protein content. The morphological features of chickpea seeds, such as colour and texture, are observable and play a major role in classifying different chickpea varieties. This process is often carried out by human experts, and is time-consuming, inaccurate, and expensive. The objective of the study was to design an automated chickpea classifier using an RGB-colour-image-based model for considering the morphological features of chickpea seed. As part of the data acquisition process, five hundred and fifty images were collected per variety for four varieties of chickpea (CDC-Alma, CDC-Consul, CDC-Cory, and CDC-Orion) using an industrial RGB camera and a mobile phone camera. Three CNN-based models such as NasNet-A (mobile), MobileNetV3 (small), and EfficientNetB0 were evaluated using a transfer-learning-based approach. The classification accuracy was 97%, 99%, and 98% for NasNet-A (mobile), MobileNetV3 (small), and EfficientNetB0 models, respectively. The MobileNetV3 model was used for further deployment on an Android mobile and Raspberry Pi 4 devices based on its higher accuracy and light-weight architecture. The classification accuracy for the four chickpea varieties was 100% while the MobileNetV3 model was deployed on both Android mobile and Raspberry Pi 4 platforms.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: GABA and Rice Research · Advanced Chemical Sensor Technologies · Spectroscopy and Chemometric Analyses
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Engineering, University of Guelph, Guelph, ON N1G 2W1, Canada
ICAR-Central Institute of Post-Harvest Engineering and Technology (CIPHET), Ludhiana 141 004, India
Funding
CARE-AI, University of Guelph · Indian Council of Agricultural Research (ICAR), India
References (40)
Show source-provided references
1. Food and Agriculture Organization (FAO) (2020). FAOSTAT Statistical Database of the United Nation Food and Agriculture Organization (FAO) Statistical Division, FAO.
2. Pulses: An overview10.1007/s13197-017-2537-4 ↗
3. Pulses and Legumes: Nutritional Opportunities and Challenges
4. Pulse proteins: Secondary structure, functionality and applications
5. Application of near-infrared hyperspectral imaging coupled with chemometrics for rapid and non-destructive prediction of protein content in single chickpea seed10.1016/j.jfca.2022.104938 ↗
6. Wood, J.A., and Grusak, M.A. (2007). Chickpea Breeding and Management, CABI.
7. Tiwari, U., and Bawa, A.S. (2007). Pulse Foods, Academic Press.
8. Canadian Grain Commission (2022, May 20). Pulse Crops Methods and Tests. Available online: https://www.grainscanada.gc.ca/en/grain-research/export-quality/pulses/methods-tests.html.
9. Chickpea varietal classification using deep convolutional neural networks with transfer learning10.1111/jfpe.13975 ↗
10. Perspectives for remote sensing with unmanned aerial vehicles in precision agriculture10.1016/j.tplants.2018.11.007 ↗
11. Application of machine vision system in food detection10.3389/fnut.2022.888245 ↗
12. Machine learning techniques for analysis of hyperspectral images to determine quality of food products: A review10.1016/j.crfs.2021.01.002 ↗
13. Quantitative reconstruction of defects in multi-layered bonded composites using fully convolutional network-based ultrasonic inversion10.1016/j.jsv.2022.117418 ↗
14. Jiang, J., Zhang, K., and Timofte, R. (2021, January 11–17). Towards flexible blind JPEG artifacts removal. Proceedings of the IEEE/CVF International Conference on Computer Vision, Montreal, BC, Canada.10.1109/ICCV48922.2021.00495 ↗
15. Quantitative detection of metanil yellow adulteration in chickpea flour using line-scan near-infrared hyperspectral imaging with partial least square regression and one-dimensional convolutional neural network10.1016/j.jfca.2023.105290 ↗
16. Economical thermal-RGB imaging system for monitoring agricultural crops10.1016/j.compag.2018.02.018 ↗
17. An analysis on the dimensions of information security culture concept: A review
18. A Creative IoT agriculture platform for cloud fog computing
19. mySense: A comprehensive data management environment to improve precision agriculture practices10.1016/j.compag.2019.05.028 ↗
20. AOAC (2005). Official Methods of Analysis, 18th ed, Association of Official Analytical Chemists.
21. A survey on image data augmentation for deep learning10.1186/s40537-019-0197-0 ↗
22. Perez, L., and Wang, J. (2017). The effectiveness of data augmentation in image classification using deep learning. arXiv.
23. Transfer learning-based deep ensemble neural network for plant leaf disease detection10.1007/s41348-021-00465-8 ↗
24. Tan, M., and Le, Q. (2019, January 9–15). Efficientnet: Rethinking model scaling for convolutional neural networks. Proceedings of the International Conference on Machine Learning, PMLR, Long Beach, CA, USA.
25. DeepMaker: A multi-objective optimization framework for deep neural networks in embedded systems10.1016/j.micpro.2020.102989 ↗
26. Young, S.R., Rose, D.C., Karnowski, T.P., Lim, S.H., and Patton, R.M. (2015, January 15–20). Optimizing deep learning hyper-parameters through an evolutionary algorithm. Proceedings of the Workshop on Machine Learning in High-Performance Computing Environments, Austin, TX, USA.10.1145/2834892.2834896 ↗
27. Saxen, F., Werner, P., Handrich, S., Othman, E., Dinges, L., and Al-Hamadi, A. (2019, January 23–25). Face attribute detection with mobilenetv2 and nasnet-mobile. Proceedings of the 2019 11th International Symposium on Image and Signal Processing and Analysis (ISPA), Dubrovnik, Croatia.10.1109/ISPA.2019.8868585 ↗
28. Koonce, B., and Koonce, B. (2021). MobileNetV3. Convolutional Neural Networks with Swift for Tensorflow: Image Recognition and Dataset Categorization, Apress.10.1007/978-1-4842-6168-2 ↗
29. Empirical analysis of a fine-tuned deep convolutional model in classifying and detecting malaria parasites from blood smears
30. Generalization performance of multi-pass stochastic gradient descent with convex loss functions
31. Identifying crop water stress using deep learning models10.1007/s00521-020-05325-4 ↗
32. Application of deep learning in food: A review10.1111/1541-4337.12492 ↗
33. Disease detection in tomato leaves via CNN with lightweight architectures implemented in Raspberry Pi 410.1016/j.compag.2020.105951 ↗
34. Oil palm fresh fruit bunch ripeness classification on mobile devices using deep learning approaches10.1016/j.compag.2021.106359 ↗
35. Ye, J., Li, X., Zhang, X., Zhang, Q., and Chen, W. (2020). Deep learning-based human activity real-time recognition for pedestrian navigation. Sensors, 20.10.3390/s20092574 ↗
36. Design and implementation of a convolutional neural network on an edge computing smartphone for human activity recognition10.1109/ACCESS.2019.2941836 ↗
37. Jain, S., Vaibhav, A., and Goyal, L. (2014, January 6–8). Raspberry Pi based interactive home automation system through E-mail. Proceedings of the 2014 International Conference on Reliability Optimization and Information Technology (ICROIT), Faridabad, India.10.1109/ICROIT.2014.6798330 ↗
38. Abd Elaziz, M., Dahou, A., Alsaleh, N.A., Elsheikh, A.H., Saba, A.I., and Ahmadein, M. (2021). Boosting COVID-19 image classification using MobileNetV3 and aquila optimizer algorithm. Entropy, 23.10.3390/e23111383 ↗
39. Transfer learning using a multi-scale and multi-network ensemble for skin lesion classification10.1016/j.cmpb.2020.105475 ↗
40. A CNN-based lightweight ensemble model for detecting defective carrots10.1016/j.biosystemseng.2021.06.008 ↗