Human Motion Retrieval Using Video or Drawn Sketch

Abstract

The importance of motion retrieval is increasing now a days. The majority of existing motion retrieval labor intensive, there has been a recent paradigm move in the animation industry with an increasing use of pre-recorded movement of animating exclusive figures. An essential need to use motion catch data is an efficient method for listing and accessing movements. I n this work, a novel sketching interface for interpreting the problem is provided. This simple strategy allows the user to determine the necessary movement by drawing several movement swings over a attracted personality, which needs less effort and extends the users expressiveness. To support the real-time interface, a specific development of the movements and the hand-drawn question is needed. Here we are implementing the Conjugate Gradient method for retrieving motion from hand drawn sketch and video. It is an optimization and prominent iterative method. It is fast and uses a small amount of storage.