Improving Classification Performance In Distorted Hyperspectral Data Using Dt-Cwt Image Fusion

This paper presents a novel method of fusing of hyperspectral bands in hyperspectral images subject to distortions typical for visual sensor networks environment. The proposed fusion method uses the Structural Similarity Measure (SSIM) to measure a level of distortion in hyperspectral bands of a received hyperspectral image in order to optimize the classification performance. The region-based image fusion algorithm using the Dual-Tree Complex Wavelet transform (DT-CWT) is employed to fuse the distorted bands with the bands most similar to distorted ones in terms of SSIM. The performance of the proposed method was tested for AVIRIS hyperspectral dataset and it improved the classification accuracy in the dataset for up to 5%.