Abstract
This paper proposes a noise reduction algorithm for images of medical X-ray fluoroscopy. Fluoroscopy generally operates at a low dose condition to reduce radiation exposures for technicians and patients. However, image noises generated by that condition degrade visibility of target objects in images. A conventional method generates image lags in the neighborhood of moving objects and another method is difficult to perform all calculations in real time. To address these problems, this paper proposes a noise reduction method, which detects motion of images using whole image information and applies time-space filter while tracking the motion. The method achieved reducing noise without image lags and real-time processing at 30 frames per seconds.
