Remote sensing image classification exploiting multiple kernel learning

TL;DR

This paper introduces a novel multiple kernel learning approach for land use classification from remote sensing images, effectively combining features without prior heuristics, especially useful with small training datasets.

Contribution

The paper presents a new MKL-based method that automatically optimizes feature combinations for remote sensing classification, improving performance with limited training data.

Findings

Effective classification on public datasets

Automatic feature combination without heuristics

Good performance with small training sets

Abstract

We propose a strategy for land use classification which exploits Multiple Kernel Learning (MKL) to automatically determine a suitable combination of a set of features without requiring any heuristic knowledge about the classification task. We present a novel procedure that allows MKL to achieve good performance in the case of small training sets. Experimental results on publicly available datasets demonstrate the feasibility of the proposed approach.