The soil classification task has become a major challenge in the field of computer vision due to subtle differences between categories, scarce samples, and traditional classification methods that are time-consuming and labor-intensive. To solve this problem, this paper proposes a soil classification algorithm combining adaptive feature transformation (AFT) and dual metric module (CCE) to improve classification accuracy and generalization ability. The AFT module dynamically adjusts image features through affine transformation, enhances the model's adaptability to changes in feature distribution, and captures subtle differences. The CCE module combines the advantages of cosine similarity and Euclidean distance to improve the accuracy of feature similarity evaluation. The classification accuracy on the public data sets CUB-200-2011 and Stanford Dogs reached 87.36% and 81.50% respectively, and the classification accuracy on the soil data set Soil was 72.43%. Experimental results show that this method achieves significant performance improvement in few-shot soil classification tasks, verifying its effectiveness in complex classification problems. Future work will further explore the impact of local feature enhancement strategies on model performance to continuously optimize the classification effect.

Hao Liang, Jiawei Wang, Ling Wei. Research on few-shot soil classification algorithm based on feature transformation and dual measurement. Academic Journal of Computing & Information Science (2024), Vol. 7, Issue 10: 89-96. https://doi.org/10.25236/AJCIS.2024.071013.

[1] Hou D, Bolan N S, Tsang D C W, et al. Sustainable soil use and management: An interdisciplinary and systematic approach[J]. Science of the Total Environment, 2020, 729: 138961.

[2] Wang D H, Zhou W, Li J, et al. Exploring misclassification information for fine-grained image classification[J]. Sensors, 2021, 21(12): 4176.

[3] Liu W, Tong L, Sun Y, et al. Automatic soil classification method from CPTU data based on convolutional neural networks[J]. Bulletin of Engineering Geology and the Environment, 2024, 83(8): 1-20.

[4] Alshahrani H, Alkahtani H K, Mahmood K, et al. Chaotic Jaya optimization algorithm with computer vision-based soil type classification for smart farming[J]. IEEE Access, 2023, 11: 65849-65857.

[5] Wang H, Frank E, Pfahringer B, et al. Feature extractor stacking for cross-domain few-shot learning[J]. Machine Learning, 2024, 113(1): 121-158.

[6] Sa L, Yu C, Ma X, et al. Attentive fine-grained recognition for cross-domain few-shot classification[J]. Neural Computing and Applications, 2022, 34(6): 4733-4746.

[7] Zhong J, Gu K, Jiang H, et al. A fine-tuning prototypical network for few-shot cross-domain fault diagnosis [J]. Measurement Science and Technology, 2024, 35(11): 116124.

[8] Zangana H M, Mohammed A K, Mustafa F M. Advancements and Applications of Convolutional Neural Networks in Image Analysis: A Comprehensive Review[J]. Jurnal Ilmiah Computer Science, 2024, 3(1): 16-29.

[9] Gao C, Li W, He L, et al. A distance and cosine similarity-based fitness evaluation mechanism for large-scale many-objective optimization[J]. Engineering Applications of Artificial Intelligence, 2024, 133: 108127.

[10] Dharsini S V, Razak M A, Modi S, et al. Captioning based image using Euclidean distance and resNet-50[C]//2022 International Conference on Data Science, Agents & Artificial Intelligence (ICDSAAI). IEEE, 2022, 1: 1-5.