Diffpurifier: An Optical and SAR Image Change Detection Method Based on Diffusion Purification

Disasters occur continuously around the world, causing significant damage to human life and property. In practice, there are inevitably situations that require the use of heterogeneous remote sensing images, for instance, optical and synthetic aperture radar (SAR) images, for change detection (CD) and disaster recognition. However, the significant differences between heterogeneous remote sensing images make direct comparison difficult. Although some methods use image translation (IT) to reduce these differences, most employ a two-stage “translation followed by CD” strategy, which can lead to feature degradation. Furthermore, these methods often require adversarial training or complex loss functions that are sensitive to hyperparameters. This article proposes a CD network for optical and SAR images, named Diffpurifier. First, optical images are translated into SAR images using pre-trained denoising diffusion probabilistic models (DDPMs) and ordinary differential equations (ODEs) while simultaneously extracting multiscale features. Then, CD is performed under superpixel enhancement to improve the homogeneity of the CD maps. Diffpurifier not only integrates IT and feature extraction, simplifying the workflow, but also maintains high accuracy, stable training, and generalization to different types of data without the need for additional translation constraints. In comparative experiments on four public datasets, Diffpurifier outperforms the second-best method by an average of approximately 5% in terms of $F1$ -score, validating the effectiveness and robustness of the method.