Super-Resolution Erlangen (SupER)数据库
收藏资源简介:
Super-Resolution Erlangen (SupER)数据库是由德国弗里德里希-亚历山大大学(FAU)埃尔兰根-纽伦堡的研究团队创建的公开可用数据库,包含超过20,000张真实低分辨率图像及其对应的高分辨率真实数据。该数据集覆盖了14个场景,通过相机硬件二值化处理,涵盖了四种空间分辨率级别。创建过程中,使用单一相机捕捉多分辨率图像序列,并通过帧平均获得高分辨率图像。SupER数据库旨在为超分辨率(SR)技术提供一个全面的基准,特别是在解决实际条件下的SR问题,如独立对象和相机运动或光度变化。
The Super-Resolution Erlangen (SupER) database is a publicly available dataset created by the research team from Friedrich-Alexander University Erlangen-Nuremberg (FAU), Germany. It contains over 20,000 real low-resolution images and their corresponding high-resolution ground-truth data. The dataset covers 14 scenarios and encompasses four spatial resolution levels generated via camera hardware binarization processing. During its construction, a single camera was employed to capture multi-resolution image sequences, and high-resolution images were obtained through frame averaging. The SupER database aims to provide a comprehensive benchmark for super-resolution (SR) technologies, especially for solving SR problems under real-world conditions such as independent object and camera motion, as well as photometric variations.




