Machine Vision and Video Analytics Solution
For a Medical Research Organization
Efficient tracking of both fiducial markers and boluses during feeding and drinking and to characterize behavioural states from high speed biplanar video fluoroscopy.
More efficient and robust algorithm for marker detection and tracking to cut the digitization time through 3D characterization of jaw and tongue movements during unconstrained feeding.
Reduced digitization time – at least 100-500 times lesser when compared to manual counterparts.
Scope of Engagement
3 Member multi-disciplinary team / 3 months.
Development of Machine Learning & Deep Learning Algorithms, for Video analytics and Marker-less movement detection.