Automatic generation of human models for motion capture, biomechanics and animation
View Patent ↗An automated method for the generation of (i) human models comprehensive of shape and joint centers information and/or (ii) subject specific models from multiple video streams is provided. To achieve these objectives, a kinematic model is learnt space from a training data set. The training data set includes kinematic models associated with corresponding morphological models. A shape model is identified as well as one or more poses of the subject. The learnt kinematic model space and the identified shape model are combined to generate a full body model of the subject starting from as few as one-static pose. Further, to generate a full body model of an arbitrary human subject, the learnt kinematic model space and the identified shape model are combined using a parameter set. The invention is applicable for fully automatic markerless motion capture and generation of complete human models.
1. A method for generating subject-specific human motion capture models having morphological and kinematic model information, comprising:
(a) learning a kinematic model space from a training data set, said training data set comprises kinematic models associated with corresponding morphological models;
(b) using a series of meshes describing human body shape to perform a shape registration algorithm for identifying and registering from a database a shape model of a human subject in one or more poses;
(c) combining said learnt kinematic model space and said identified shape model to generate a full body model of said human subject, said generating starts from as few as one of said poses identified in step (b); and
(d) combining said learnt kinematic model space and said identified shape model to generate a full body model of an arbitrary human subject, said generating using a parameter set.
2. The method as set forth in claim 1 , wherein said step of combining in 1(c) and/or 1(d) are based on matching to a three-dimensional representation of the subject.
3. The method as set forth in claim 1 , wherein said kinematic models define an articulated multi-segmented rigid body model.
4. The method as set forth in claim 1 , wherein said steps 1(a) through 1(d) are performed in a fully automated fashion.
5. The method as set forth in claim 1 , wherein said step 1(d) is performed in an automated fashion.
6. The method as set forth in claim 1 , wherein said step of combining in 1(c) and/or 1(d) further comprises imposing joint constraints.