Deep Self-taught Learning for Remote Sensing Image Classification

TL;DR

This paper introduces a deep self-taught learning framework that enhances remote sensing image classification by learning hierarchical feature representations, leading to improved accuracy over traditional methods.

Contribution

It presents a novel deep self-taught learning approach utilizing sparse representation and dictionary learning for remote sensing classification.

Findings

Deep features improve classification accuracy

Framework effective on multispectral Landsat data

Outperforms traditional feature-based methods

Abstract

This paper addresses the land cover classification task for remote sensing images by deep self-taught learning. Our self-taught learning approach learns suitable feature representations of the input data using sparse representation and undercomplete dictionary learning. We propose a deep learning framework which extracts representations in multiple layers and use the output of the deepest layer as input to a classification algorithm. We evaluate our approach using a multispectral Landsat 5 TM image of a study area in the North of Novo Progresso (South America) and the Zurich Summer Data Set provided by the University of Zurich. Experiments indicate that features learned by a deep self-taught learning framework can be used for classification and improve the results compared to classification results using the original feature representation.