DDCNet: A Dual-Domain Collaborative Network for Remote Sensing Image Dehazing
Qilin LIN, Hao ZHOU, Weiwei Jiang · IEICE Transactions on Information and Systems · 2026
Remote sensing image dehazing is challenging due to complex scenes and uneven haze, while spatial-based methods struggle to recover high-frequency details. To address these issues, we propose a dual-domain collaborative dehazing network that integrates spatial-domain representation with frequency-domain enhancement. Specifically, the Discrete Wavelet Transform is employed to decompose features into low-frequency and high-frequency components, enabling targeted restoration. A Spatial Processing Module is designed to capture large-scale contextual dependencies, while a High-Frequency Restoration Module focuses on recovering texture and edge information in remote sensing imagery. Furthermore, a Cross-Domain Interaction Module is introduced to facilitate effective information exchange between spatial and frequency domains. Extensive experiments demonstrate that the proposed method achieves superior performance in terms of structural preservation and detail recovery.