Polarimetric Imaging Scenes – Over 100 Scenarios

This dataset is a comprehensive polarized image dataset designed to support research in remote sensing, polarization analysis, and super-resolution imaging. It comprises hundreds of distinct scenes, each captured using a custom-built DOT-based polarization acquisition system integrated with a standard camera. For every scene, four images corresponding to different polarization channels are provided. These multi-channel polarized images contain rich textures and fine-grained structural details, covering diverse content such as buildings, trees, and distant targets—making the dataset particularly valuable for studies focused on complex natural and man-made environments. From these polarization channel images, polarization parameters can be derived, enabling deeper analysis in polarization-based remote sensing. Furthermore, this dataset is well-suited for training and evaluating deep learning models in tasks such as polarization image super-resolution, target recognition, and scene understanding. The variety of scene types and the precision of the polarized data make this resource highly beneficial for advancing algorithms in both image reconstruction and interpretation within the context of polarimetric remote sensing.