DFG-DDM: Deep Frequency-Guided Denoising Diffusion Model for Remote Sensing Image Dehazing

Haze removal in remote sensing (RS) images has become increasingly vital due to their capacity to contain essential information for accurate geospatial analysis. Notably, this phenomenon is particularly pronounced in both spatial and spectral distributions of buildings, complex terrain, and landforms. Inspired by the success of generative models in enhancing details incrementally and suppressing noise, we propose a deep frequency-guided denoising diffusion model (DFG-DDM) for RS imagery dehazing. The pixel-level generative capability of the diffusion model is fully leveraged, and the fast Fourier transform (FFT) is utilized to extract frequency-domain information. This enables the separate mining of semantic information from RS images in both spatial and spectral domains. Concurrently, the continuity of the image in the frequency domain is ensured without altering the diffusion process, thus achieving detailed retention while improving overall clarity. Furthermore, to address the scarcity of physically realistic training data for spatially heterogeneous atmospheric degradation, we construct a random haze distribution dataset for remote sensing (RHDRS) dehazing. RHDRS randomly simulates the spatial distribution and thickness of haze, containing 4500 hazy images along with the corresponding ground truths (GTs). Experiments demonstrate that our approach outperforms existing state-of-the-art techniques. The dataset and the code can be accessed at https://github.com/Junjie-LLL/DFG-DDM