Volume V | Issue 2 | 2026 | Paper ID: IJATEM-V05I02P2

Enhanced HSI classification using depth-wise separable CNN with advanced feature extraction and selection techniques

Author(s)

Yallamandaiah. S, M. Sanjana, N. Sindhu Bhargavi, E. Brahmani, S. Anjali, T. Pravallika

Cite This

Yallamandaiah. S, M. Sanjana, N. Sindhu Bhargavi, E. Brahmani, S. Anjali, T. Pravallika, "Enhanced HSI classification using depth-wise separable CNN with advanced feature extraction and selection techniques", International Journal of Advanced Trends in Engineering and Management, vol. 05, no. 02, pp.16-24, 2026.

Abstract

Hyperspectral Image classification has many challenges. The challenges are caused due to the high dimensionality of the data and variability. The results in lower accuracy. The Existing methods cannot handle the high dimensional data, and poor feature selection and gives lower accuracy. To solve these issues this research addresses the ResNet50 model is used to extract the important features. IARO algorithm (Artificial Rabbits Optimization) is used to select the best features. The Classification is done using the Depth-wise Separable Convolutional Network. It is used to enhance the performance of the proposed method. The proposed model improves the quality of data and increases the accuracy of the model. This addresses the limitations of the traditional methods, and this proposed method offers more reliability and effective classification system. Overall, method performed better than many other existing approaches. It improved accuracy and also reduced the complexity of the process.

Keywords

Hyperspectral Image Classification, ResNet50, Feature Extraction, Artificial Rabbits Optimization (IARO), Depth-wise Separable CNN

References

[1] Li Z, Zheng K, Li J, Li C, Gao L (2024) Cross semantic heterogeneous modeling network for hyperspectral image classification. IEEE Trans GeosciRemoteSens.

CrossRef Google Scholar Publisher Link
[2] Xu J, Li K, Li Z, Chong Q, Xing H, Xing Q,Ni M(2024) Fuzzy graph convolutional network for hyperspectral image classification. Eng Appl Artif Intell 127:107280.

CrossRef Google Scholar Publisher Link
[3] Jiang M, Su Y, Gao L, Plaza A, Zhao XL, Sun X, Liu G (2024) GraphGST: Graph generative structure-aware transformer for hyperspectral image classification. IEEETransGeosciRemoteSens.

CrossRef Google Scholar Publisher Link
[4] Ahmad M, Usama M, Khan AM, Distefano S, Altuwaijri HA, Mazzara M (2024) Spatial spectral transformer with conditional position encoding for hyperspectral image classification. IEEE GeosciRemoteSensLett.

CrossRef Google Scholar Publisher Link
[5] Huang Y, Peng J, Zhang G, Sun W, Chen N, Du Q (2024) Adversarial domain adaptation network with calibrated prototype and dynamic instance convolution for hyperspectral image classification. IEEE Trans Geosci Remote Sens.

CrossRef Google Scholar Publisher Link
[6] Lv H, Li Y, Zhang H, Wang R (2024) Multi-dimensional deep dense residual networks and multiple kernel learning for hyper spectral image classification. Infrared Phys Technol 138:105265.

CrossRef Google Scholar Publisher Link
[7] Huang S, Liu Z, Jin W, Mu Y (2024) Superpixel-based multi scale multi-instance learning for hyperspectral image classification. PatternRecogn149:110257.

CrossRef Google Scholar Publisher Link
[8] Zhang Z, Zhou H, Zhang C, Zhang X, Jiang Y (2023) A multi-level deformable gated aggregated network for hyperspectral image classification. Int J Appl Earth Obs Geoinf 123:103482.

CrossRef Google Scholar Publisher Link
[9] Li Z, Meng Q, Guo F, Wang L, Huang W, Hu Y, Liang J (2023) Feature-guided dynamic graph convolutional network for wet land hyperspectral image classification. Int J Appl Earth Obs Geoinf123:103485.

CrossRef Google Scholar Publisher Link
[10] Xin Z, Li Z, Xu M, Wang L, Ren G, WangJ, HuY (2024) Feature disentanglement-based domain adaptation network for cross-scene coastal wetland hyperspectral image classification. Int J Appl Earth Obs Geoinf129:103850.

CrossRef Google Scholar Publisher Link
[11] Liang M, Zhang X, Yu X, Yu L, Meng Z, Zhang X, Jiao L (2024) An efficient Transformer with neighborhood contrastive tokenization for hyperspectral images classification. Int J Appl EarthObsGeoinf131:103979.

CrossRef Google Scholar Publisher Link
[12] Kavitha K, Arivazhagan S, Suriya B (2014) Classification of Pavia University hyperspectral image using Gabor and SVM classifier. Int J New Trends Electron Commun 2(3): 9–14.

Google Scholar Publisher Link
[13] Liu H, Li W, Xia X G, Zhang M, Tao R (2023) Multiarea target attention for hyperspectral image classification. IEEE Trans Geosci Remote Sens 61:

CrossRef Google Scholar Publisher Link
[14] Luo X, Tong X, Pan H (2020) Integrating multi-resolution and multitemporal Sentinel-2 imagery for land-cover mapping in the Xiongan New Area, China. IEEE Trans Geosci Remote Sens 59(2):10291040.

CrossRef Google Scholar Publisher Link

Full Length Article PDF

Article Details