A Deep Learning Denoising Framework Based on FFDNet for SAR Image Change Detection

Synthetic aperture radar (SAR) image change detection is an important research direction in remote sensing field. Due to the influence of multiplicative noise, it is difficult to accurately obtain the differential images of SAR images at different times. We proposed a FFDNet-based deep learning image change detection framework, which achieved a good tradeoff between inference speed and denoising performance. In addition, we also designed a bidirectional connected domain processing algorithm to further improve the SAR image change detection process. The experimental results of several data sets show that the framework has high accuracy and efficiency, which can be used as a reference for the research in related fields.