Image Processing for Motion Magnification

Abstract

Motion Magnification (MM) is a collection of relatively recent techniques within the realm of Image Processing. The main motivation of introducing these techniques is to support the human visual system to capture relevant displacements of an object of interest; these motions can be in object color and in object location. In fact, the goal is to opportunely process a video sequence to obtain as output a new video in which motions are magnified and visible to the viewer. We propose a numerical technique based on the Phase-Based approach for Motion Magnification which analyses the video sequence in the Fourier Domain and rely on the Fourier Shifting Property. We describe the mathematical basis of this method and the corresponding implementation in a numerical algorithm. We present preliminary experiments, focusing on some basic examples made up using synthetic images.