It is constructed using the incarnation of depth camera

摘要

A kind of building avatar-based humans include the object for obtaining depth map, obtain dummy skeleton, the depth map based on subject and have gathered in measurement from one group of characteristic of dummy skeleton. This measurement corresponds to the dummy skeleton of the distance between scheduled point. In this illustrative methods, which is provided as input to algorithm training machine learning areas. The algorithm can be in the use scope of trained manikin posture and the single posture of people's model, and exports the measurement of the functional characteristic of virtual body net. This method also includes that construction effective head grid is different from virtual body net, is similar to those objects, connection virtual subject net and effective head grid with facial features.