Motion detection in infrared retinal image sequences

Infrared image data captured by non-mydriatic digital retinography systems often are used in the diagnosis and treatment of the age-related macular disease. Infrared illumination is less aggressive to the patient retina, and retinal studies can be carried out without pupil dilation. However, sequences of infrared eye fundus images of static scenes, tend to present pixel intensity fluctuations in time, and noisy and background illumination changes pose a challenge to most motion detection methods proposed in the literature. In this paper, we present a new retina motion detection method that is adaptive to background noise and illumination changes. Our preliminary experimental results indicate that this method potentially is suitable for detecting retinal motion in infrared image sequences, and could be used in laser treatment systems for the age-related macular disease.