MASCA: Multiscale Adaptive Spectral Component Analysis for Hyperspectral Image Classification
- 7 May 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 23 (1545598X) , 5504405
Abstract
In this paper, we propose Multi-Scale Adaptive Spectral Component Analysis (MASCA), a frequency-domain preprocessing method that adaptively identifies discriminative Discrete Cosine Transform (DCT) components through multi-scale peak analysis. MASCA assigns a relevance score to each DCT component by detecting magnitude transitions across multiple scales while integrating spatial coherence and spectral stability constraints, with hyperparameters tuned via Bayesian optimization. Experimental results on four benchmark hyperspectral datasets (Indian Pines, Pavia University, Botswana, and Salinas) demonstrate the effectiveness of the proposed method.Keywords
Funding Information
- “Région Bretagne” through the 2023 Regional Call for Projects “L’innovation collaborative au croisement des filières”
This publication has 17 references indexed in Scilit:
- CasFormer: Cascaded transformers for fusion-aware computational hyperspectral imagingInformation Fusion, 2024
- Spectral–Spatial Self-Attention Networks for Hyperspectral Image ClassificationIEEE Transactions on Geoscience and Remote Sensing, 2021
- An overview on spectral and spatial information fusion for hyperspectral image classification: Current trends and challengesInformation Fusion, 2020
- DCT-Based Local Descriptor for Robust Matching and Feature Tracking in Wide Area Motion ImageryIEEE Geoscience and Remote Sensing Letters, 2020
- Feature Extraction for Hyperspectral Imagery: The Evolution From Shallow to Deep: Overview and ToolboxIEEE Geoscience and Remote Sensing Magazine, 2020
- Deep learning classifiers for hyperspectral imaging: A reviewISPRS Journal of Photogrammetry and Remote Sensing, 2019
- Random ForestsMachine Learning, 2001
- Independent Component Analysis2001
- Learning the parts of objects by non-negative matrix factorizationNature, 1999
- Support-vector networksMachine Learning, 1995