Edge-aware spatial filtering-based motion magnification

In this paper, we present a novel approach toward motion magnification using edge-aware filtering. Edge-aware filters have been used for applications such as image denoising, enhancement, and tone mapping. In this work, we have used fast local Laplacian filter followed by Eulerian motion magnification for enhancement of subtle changes in the video. The proposed method aims to reduce the noise and artifacts in the output video apart from motion magnification. Experiments are conducted on sample videos having minute changes which are imperceptible to see with the human eyes. A no-reference video quality assessment has been employed using a video quality score in order to quantify the motion magnification results.