Video Processing with Eulerian Video Magnification
Our objective is to reveal subtle temporal changes in videos that are difficult or impossible to detect with the naked eye. Our method, called Eulerian Video Magnification, processes standard video sequences through spatial decomposition followed by temporal filtering of frames. The resulting signals are then amplified to uncover hidden information. Using this technique, we can visualize blood flow patterns as they propagate through facial tissue and magnify minute movements. Our implementation can operate in real-time, displaying phenomena at user-selectable temporal frequencies with appropriate signal processing techniques.