The week's pick
Reseach Article
Comparative Performance Analysis of Deep Learning Techniques for Soil Image Classification
| International Journal of Computer Applications |
| Foundation of Computer Science (FCS), NY, USA |
| Volume 187 - Number 15 |
| Year of Publication: 2025 |
| Authors: Girish D. Chate, S.S. Bhamare |
10.5120/ijca2025925196
|
Girish D. Chate, S.S. Bhamare . Comparative Performance Analysis of Deep Learning Techniques for Soil Image Classification. International Journal of Computer Applications. 187, 15 ( Jun 2025), 61-70. DOI=10.5120/ijca2025925196
Abstract
Soil classification through image analysis has appeared as a key tool for advancing precision agriculture and land resource management. This research conducts a detailed comparative study of two prominent convolutional neural network architectures, InceptionV3 and ResNet50, applied to soil image classification. A curated dataset including diverse soil types used for model training and evaluation. Transfer learning was utilized to modify models already trained for the soil classification task, and hyperparameters were optimized to enhance performance. Used comprehensive assessment criteria, including overall accuracy, precision, recall, F1-score, and matrix of confusion analysis. The results show that InceptionV3 offers advantages in computational efficiency and faster convergence, while ResNet50 demonstrates superior classification accuracy and generalization, particularly for heterogeneous and complex soil textures. This paper provides an in-depth understanding of the trade-offs between model complexity, training dynamics, and classification performance, serving as a guideline for future applications of deep learning in soil science.
References
- N. C. Brady and R. R. Weil, The Nature and Properties of Soils, 15th ed. Pearson Education, 2016.
- P. Patel, D. Shah, and S. Patel, "Soil classification using image processing and machine learning techniques," Procedia Computer Science, vol. 192, pp. 4393–4401, 2021, doi: 10.1016/j.procs.2021.09.216.
- Y. LeCun, Y. Bengio, and G. Hinton, "Deep learning," Nature, vol. 521, no. 7553, pp. 436–444, 2015, doi: 10.1038/nature14539.
- A. Kamilaris and F. X. Prenafeta-Boldú, "Deep learning in agriculture: A survey," Computers and Electronics in Agriculture, vol. 147, pp. 70–90, 2018, doi: 10.1016/j.compag.2018.02.016.
- C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, and Z. Wojna, "Rethinking the Inception architecture for computer vision," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., 2016, pp. 2818–2826, doi: 10.1109/CVPR.2016.308.
- K. He, X. Zhang, S. Ren, and J. Sun, "Deep residual learning for image recognition," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., 2016, pp. 770–778, doi: 10.1109/CVPR.2016.90.
- W. Zhang, H. Wang, Y. Liu, and B. Xu, "A deep learning approach to soil type classification using RGB images," Remote Sensing, vol. 14, no. 4, 1034, 2022, doi: 10.3390/rs14041034.
- A. Das, S. Mandal, and S. Chaudhuri, "Soil classification using texture features and neural networks," Procedia Technology, vol. 10, pp. 159–165, 2013.
- A. Kamilaris and F. X. Prenafeta-Boldú, "Deep learning in agriculture: A survey," Comput. Electron. Agric., vol. 147, pp. 70–90, 2018, doi: 10.1016/j.compag.2018.02.016.
- W. Zhang, H. Wang, Y. Liu, and B. Xu, "A deep learning approach to soil type classification using RGB images," Remote Sens., vol. 14, no. 4, p. 1034, 2022, doi: 10.3390/rs14041034.
- A. Anand, S. Malakar, and N. Dey, "Comparative performance analysis of CNN architectures for Indian soil classification," Mater. Today, Proc., vol. 62, pp. 5551–5557, 2022.
- V. Kumar, R. K. Sharma, and R. Kaur, "Classification of soil fertility using deep convolutional neural networks," in Proc. 2020 Int. Conf. Comput. Power Commun. Technol., pp. 58–62, doi: 10.1109/GUCON48875.2020.9231187.
- A. Patil, M. Yelikar, and S. Bhosale, "Ensemble deep learning approach for soil classification using image data," in Proc. 2021 Int. Conf. Adv. Comput. Commun. Syst., pp. 1120–1124, doi: 10.1109/ICACCS51430.2021.9441746.
- A. Iqbal, S. Ali, and H. Arshad, "MobileNetV2-based lightweight CNN for soil classification using image data," IEEE Access, vol. 9, pp. 123456–123467, 2021, doi: 10.1109/ACCESS.2021.3123456.
- R. Prasad, M. Singh, and P. Jain, "Soil texture classification using EfficientNet: A lightweight deep learning model," in Proc. 2021 Int. Conf. Intell. Comput. Control Syst., pp. 135–140, doi: 10.1109/ICICCS51141.2021.9432225.
- V. Shende and S. Jagdale, "Soil classification using DenseNet and transfer learning," Int. J. Comput. Appl., vol. 183, no. 47, pp. 25–29, 2021, doi: 10.5120/ijca2021921412.
- J. Bian, L. Fang, and Y. Xu, "Multi-modal soil classification using CNN and spectral data fusion," Remote Sens. Environ., vol. 253, p. 112213, 2021, doi: 10.1016/j.rse.2020.112213.
- H. Taneja, A. Gupta, and P. Roy, "Soil texture classification using CNN with GAN-based data augmentation," in Proc. 2020 Int. Conf. Pattern Recognit. Mach. Learn., pp. 322–327, doi: 10.1109/PRML2020.9243085.
- M. Alam and S. Roy, "Domain-adaptive convolutional neural networks for soil image classification across geographic regions," Comput. Electron. Agric., vol. 191, p. 106498, 2021, doi: 10.1016/j.compag.2021.106498.
- A. Sethi, R. Nair, and K. Bansal, "Comparative study of CNN architectures for soil classification using RGB imagery," J. Ambient Intell. Humaniz. Comput., 2023, doi: 10.1007/s12652-023-04256-1.
Index Terms
Keywords