Comparative Analysis of Deep Learning Techniques for Soil Image Classification
DOI:
https://doi.org/10.26438/ijcse/v13i4.1522Keywords:
Agriculture, Soil, Image Classification, VGG16, VGG19Abstract
Soil classification is a crucial step in agricultural and environmental planning. Current innovations in computer vision and deep learning have enabled automatic soil classification using image-based approaches. This paper, explore comparative analysis of two popular convolutional neural network architectures, VGG16 and VGG19, for soil image classification. A use of soil image dataset containing various soil types used to evaluate the performance of both models. These models fine-tuned using transfer learning, and performance determined using metrics such as accuracy, precision, recall, F1-score, and training time. The result shows that both VGG16 and VGG19 achieve high classification accuracy, with VGG19 slightly outperforming than VGG16 in terms of accuracy but requiring more computational resources and time. This paper demonstrates the effectiveness of deep learning models in soil image classification and provides understandings into their comparative performance.
References
[1] S. A. Khan, et al., “A review of soil classification using machine learning,” Geoderma, Art. no. 114984, Vol.389, 2021.
[2] R. McBratney, D. J. Field, and A. B. McBratney, “Future directions of soil science,” Geoderma Regional, Vol.4, pp.1–5, 2015.
[3] M. A. Shepherd and M. G. Shepherd, “Modern soil testing and classification,” Soil Use and Management, Vol.29, No.4, pp.515–525, 2013.
[4] R. Goswami, A. Saha, and P. Saha, “Soil classification using Convolutional Neural Networks,” Computers and Electronics in Agriculture, Vol.172, 2020.
[5] D. Yadav, R. K. Sharma, and N. Kumari, “Deep learning models for soil type classification using image and spectral data,” Environmental Modelling & Software, Vol.148, 2022.
[6] A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet classification with deep convolutional neural networks,” in Proc. NIPS, pp. 1097–1105, 2012.
[7] K. Simonyan and A. Zisserman, “Very Deep Convolutional Networks for Large-Scale Image Recognition,” arXiv:1409.1556, 2014.
[8] Srivastava P., Shukla A., & Bansal A., A comprehensive review on soil classification using deep learning and computer vision techniques. Multimedia Tools and Applications, Vol.80, Issue.3, pp.4667-4694, 2021.
[9] Deng, Z., Liu, M., & Wang, Y., Hyperspectral image classification using convolutional neural networks and 3D-CNNs. Journal of Applied Remote Sensing, Vol.14, Issue.1, pp.014520, 2020.
[10] Ramesh P., & Suruliandi A., Soil classification using deep convolutional neural networks for Indian soil datasets. International Journal of Computer Applications, Vol.178, Issue.21, pp.27-34, 2019.
[11] Zhang, Y., Li, J., & Chen, Y., Using attention mechanisms for soil classification based on CNNs: A case study in agriculture. Agricultural Systems, 190, pp.103092, 2021.
[12] Hossain, M. S., Rahman, M. M., & Goh, A. T. C., Explainable AI for soil classification: Using Grad-CAM to visualize CNN decisions. Expert Systems with Applications, 189, pp.116030, 2022.
[13] Nguyen, D. D., Tran, T. P., & Pham, H. D., Federated learning for privacy-preserving soil image classification. International Journal of Agricultural Science and Technology, Vol.13, Issue.1, pp.32-43, 2023.
[14] Alam, M. A., Islam, M. R., & Shikder, M. I., Enhancing soil image classification performance with data augmentation and image enhancement techniques. Computers and Electronics in Agriculture, 194, 106754, 2022.
[15] P. Goswami, P. P. Roy, and S. Bandyopadhyay, “Soil classification using convolutional neural networks,” J. Soil Sci. Environ. Manage., Vol.11, No.4, pp.55-63, 2020.
[16] S. A. Khan, M. Usman, and A. R. Shah, “A review of soil classification using machine learning,” Geoderma, Vol.389, pp. 114984, 2021.
[17] R. Yadav, D. Gupta, and R. Bhattacharyya, “Hybrid deep learning models for soil fertility classification using image and spectral data,” Environ. Sci. Pollut. Res., Vol.29, No.5, pp.6592-6606, 2022.
[18] S. Nandhini and V. Kavitha, “Soil texture classification using deep convolutional neural networks on satellite images,” Remote Sensing Lett., Vol.11, No.3, pp.299-308, 2020.
[19] R. Patil, M. K. Choudhury, and H. Joshi, “Real-time soil type prediction using deep learning with a mobile application,” Sensors, Vol.21, No.9, pp.2995, 2021.
[20] F. Zhao, X. Li, and Y. Zhang, “A comparative study of CNN architectures for soil profile classification using RGB images,” Agric. Informatics, Vol.3, No.2, pp.113-124, 2022.
[21] Zhang X, Youman NH, King RL. Soil texture classification using wavelet transform and maximum likelihood approach. 7929-7931
[22] Ok S, Hyun K, Youl K (2012) Texture Classification Algorithm Using RGB Characteristics of Soil Images 57, pp.393–397, 2003.
[23] Shenbagavalli R, Ramar K., Classification of soil textures based on Laws features extracted from preprocessing images on sequential and random windows. Bonfring Int J Adv Image Process 1: pp.15–18, 2011.
[24] Sofou A, Evangelopoulos G, Maragos P., Soil image segmentation and texture analysis: a computer vision approach. Geoscience and Remote Sensing Letters Vol.2, Issue.4, pp.394–398, 2005.
[25] Breul P, Gourves R., In field soil characterization: approach based on texture image analysis. Journal of Geotechnical and Geoenvironmental Engineering Vol.132, Issue.1, pp.102–107, 2006.
[26] Shenbagavalli R, Ramar K., Feature extraction of soil images for retrieval based on statistics. Int J Comput Appl 88: pp.8–12, 2014.
[27] Honawad PSK, Chinchali PSS, Pawar PK, Deshpande PP., Soil Classification and Suitable Crop Prediction. pp.25–29, 2017.
[28] Murti GSRK, Satyanarayana KVS. Influence of chemical characteristics in the development of soil colour. Geoderma 5: pp.243–248, 1971.
[29] Chung SO, Cho KH, Cho JW, Jung KY, Yamakawa T. Soil texture classification algorithm using RGB characteristics of soil images. J Fac Agric Kyushu Univ 57: pp.393–397, 2012.
[30] de O. Morais PA, Souza DM, de M. Carvalho MT, Madari BE, de Oliveira AE, Predicting soil texture using image analysis. Microchem. J. 146: pp.455–463, 2019.
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors contributing to this journal agree to publish their articles under the Creative Commons Attribution 4.0 International License, allowing third parties to share their work (copy, distribute, transmit) and to adapt it, under the condition that the authors are given credit and that in the event of reuse or distribution, the terms of this license are made clear.
