Further results on dissimilarity spaces for hyperspectral images RF-CBIR

TL;DR

This paper introduces a relevance feedback approach using dissimilarity spaces to improve hyperspectral image retrieval, addressing previous limitations in applying machine learning techniques to spectral and spatial features.

Contribution

It proposes a novel RF method based on dissimilarity spaces tailored for hyperspectral CBIR, validated on real datasets, enhancing retrieval performance.

Findings

Effective RF method improves hyperspectral CBIR accuracy

Dissimilarity space approach enables better machine learning application

Validated on real hyperspectral dataset with positive results

Abstract

Content-Based Image Retrieval (CBIR) systems are powerful search tools in image databases that have been little applied to hyperspectral images. Relevance feedback (RF) is an iterative process that uses machine learning techniques and user's feedback to improve the CBIR systems performance. We pursued to expand previous research in hyperspectral CBIR systems built on dissimilarity functions defined either on spectral and spatial features extracted by spectral unmixing techniques, or on dictionaries extracted by dictionary-based compressors. These dissimilarity functions were not suitable for direct application in common machine learning techniques. We propose to use a RF general approach based on dissimilarity spaces which is more appropriate for the application of machine learning algorithms to the hyperspectral RF-CBIR. We validate the proposed RF method for hyperspectral CBIR systems over a real hyperspectral dataset.