Extraction of temporal motor activity signals from video recordings of neonatal seizures by feature tracking methods based on deformable motion models

This paper presents a new feature tracking method for video. This method estimates the displacement of a feature between two successive frames by minimizing an error function defined in terms of the feature intensities at these frames. Feature tracking relies on an affine motion model, which can be used to track features that may be translated and deformed from one frame to the next. The proposed method is used to extract temporal motor activity signals from video recordings of neonatal seizures.