Datasets/ FIRE: Fundus Image Registration Dataset

The dataset consists of 129 retinal images forming 134 image pairs. These image pairs are split into 3 different categories depending on their characteristics. The images were acquired with a Nidek AFC-210 fundus camera, which acquires images with a resolution of 2912x2912 pixels and a FOV of 45° both in the x and y dimensions. Images were acquired at the Papageorgiou Hospital, Aristotle University of Thessaloniki, Thessaloniki from 39 patients.