Extraction of Temporal Motion Velocity Signals from Video Recordings of Neonatal Seizures by Optical Flow Methods

This paper presents a procedure developed to extract quantitative information from video recordings of neonatal seizures in the form of temporal motion velocity signals. These signals are obtained from the velocity fields computed for successive frames of the video recording by a variety of optical flow methods. The experimental results provide evidence that the motion velocity signals produced by the proposed procedure constitute an effective representation of videotaped clinical events and can be used for seizure recognition and characterization. Optical flow is the term used to indicate the velocity field generated by the relative motion between an object and the camera in a frame sequence. Optical flow provides important information for analyzing motion in video. In the absence of any additional assumptions about the nature of motion, optical flow computation based on two successive frames is an ill-posed problem. A problem is called ill-posed if its solution is not unique and/or if its solution does not depend continuously on the data [2]. Let ( , , ) I I x y t = denote the continuous space-time intensity distribution. If the intensity remains constant along a motion trajectory, then ( , , ) 0 dI x y t dt = . This latter condition can also be written as Keywords—Motion velocity signal, neonatal seizure, optical flow, velocity field, video recording