Multi-agent Deep Reinforcement Learning for Hyperspectral Feature Extraction

Hyperspectral image feature extraction plays a crucial role in reducing redundancy and correlation among spectral bands while preserving essential information. Knowledge-based feature extraction methods, such as spectral indices (SIs), leverage the interaction mechanisms between electromagnetic waves and materials to enhance the characteristic attributes of ground objects through band operations. These methods offer key advantages, including strong physical interpretability, simple construction, and robust cross-domain generalization. However, most existing SIs still rely on expert knowledge tailored to specific scenarios, leading to inherent limitations such as subjectivity, high time consumption, and implementation complexity. To address these challenges, this paper proposes a Hyperspectral Image Multi-Agent Deep Reinforcement Learning Feature Extraction algorithm (HMAFE), aiming to alleviate the burden of manual spectral index design for human experts. HMAFE employs a heuristic “generation-selection” strategy to simulate the decision-making process of domain experts. To accelerate exploration in high-dimensional action spaces, the model incorporates a multi-agent deep reinforcement learning (MADRL) framework. Experimental results demonstrate that the proposed method outperforms state-of-the-art feature selection and automated feature engineering (AutoFE) approaches in terms of both feature extraction efficiency and overall performance.