发明名称 CO-REGISTRATION - SIMULTANEOUS ALIGNMENT AND MODELING OF ARTICULATED 3D SHAPES
摘要 Present application refers to a method, a model generation unit and a computer program (product) for generating trained models (M) of moving persons, based on physically measured person scan data (S). The approach is based on a common template (T) for the respective person and on the measured person scan data (S) in different shapes and different poses. Scan data are measured with a 3D laser scanner. A generic personal model is used for co-registering a set of person scan data (S) aligning the template (T) to the set of person scans (S) while simultaneously training the generic personal model to become a trained person model (M) by constraining the generic person model to be scan-specific, person-specific and pose-specific and providing the trained model (M), based on the co registering of the measured object scan data (S).
申请公布号 US2015262405(A1) 申请公布日期 2015.09.17
申请号 US201214433178 申请日期 2012.12.14
申请人 MAX-PLANCK-GESELLSCHAFT ZUR FÖRDERUNG DER WISSENSCHAFTEN E.V. 发明人 Black Michael;Hirshberg David L.;Loper Matthew;Rachlin Eric;Weiss Alex
分类号 G06T13/40;G06T19/20;G06T17/20 主分类号 G06T13/40
代理机构 代理人
主权项 1. Model generation unit for generating trained models (M) of moving objects, based on physically measured object scan data (S), comprising: A template interface, which is adapted to provide at least one common template (T) for the respective object A sensing device for measuring object scan data (S) in different shapes and in different poses Access means to a database, storing at least one generic object model An initializing interface, which is adapted to provide initialized object scan data (S) and initialized template data (T) A Co-Registration unit, which is adapted for executing a non-linear objective function for co-registering a set of initialized object scan data (S) by executing registering and model generation in a combined manner, namely: Aligning the initialized template (T) to the set of initialized object scans (S) while simultaneously: Training the generic object model by minimizing errors in order to derive a trained object model (M) by constraining the generic object model to be scan-specific, object-specific and pose-specific An output interface for providing at least a set of trained models (M), maximally consistent with the measured object scan data (S).
地址 Munich DE