Unsupervised Band Selection Using Fused HSI and LiDAR Attention Integrating With Autoencoder

TL;DR

This work tackles unsupervised hyperspectral band selection by integrating LiDAR-derived spatial cues with HSI through a dual-attention fused mask, processed by a convolutional Autoencoder for reconstruction-based band selection. A novel distance metric combines normalized attention and Pearson-based dissimilarity, guiding hierarchical clustering to select a small, diverse, informative set of bands; the objective includes reconstruction loss and sparsity via $J(\theta)=\frac{1}{2}\| f((X_{HSI}\odot M_{fused}))-X_{HSI}\|_2^2 + \lambda \|M_{fused}\|_{2,1}$. The method is validated on Houston 2013, Trento, and MUUFL datasets, showing superior classification accuracy compared with existing unsupervised band selection and fusion models, especially when LiDAR features are incorporated. This approach yields more efficient hyperspectral analysis with fewer bands and demonstrates practical impact for remote sensing tasks under limited labeled data regimes.

Abstract

Band selection in hyperspectral imaging (HSI) is critical for optimising data processing and enhancing analytical accuracy. Traditional approaches have predominantly concentrated on analysing spectral and pixel characteristics within individual bands independently. These approaches overlook the potential benefits of integrating multiple data sources, such as Light Detection and Ranging (LiDAR), and is further challenged by the limited availability of labeled data in HSI processing, which represents a significant obstacle. To address these challenges, this paper introduces a novel unsupervised band selection framework that incorporates attention mechanisms and an Autoencoder for reconstruction-based band selection. Our methodology distinctively integrates HSI with LiDAR data through an attention score, using a convolutional Autoencoder to process the combined feature mask. This fusion effectively captures essential spatial and spectral features and reduces redundancy in hyperspectral datasets. A comprehensive comparative analysis of our innovative fused band selection approach is performed against existing unsupervised band selection and fusion models. We used data sets such as Houston 2013, Trento, and MUUFLE for our experiments. The results demonstrate that our method achieves superior classification accuracy and significantly outperforms existing models. This enhancement in HSI band selection, facilitated by the incorporation of LiDAR features, underscores the considerable advantages of integrating features from different sources.

Figures (11)

• Figure 1: Flowchart of the HSI band Selection
• Figure 2: Detailed Architecture of HSI and LiDAR MASK fusion based on Attention Module for HSI band Selection and Autoencoder Reconstruction.
• Figure 3: SVM OA Comparison-Based on UH2013 Data set using Different Band Selection Methods.
• Figure 4: KNN OA Comparison-Based on UH2013 Data set using Different Band Selection Methods.
• Figure 5: CNN OA Comparison-Based on UH2013 Data set using Different Band Selection Methods.