Human Posture Reconstruction from Monocular Images based on Criterion Function

J. Zhao, L. Li, and K.C. Keong (Singapore)


Reconstruction, Human Animation, Criterion Function


This paper presents a new approach towards the reconstruction of human posture from monocular video images that contain unrestricted human posture and human movement. It is a way towards low cost motion capture and at the same time avoids many limitations of the classical motion capture methods. A parameterized human skeleton model based on anatomy is used with the angular constraints encoded in the joints. Criterion Function is defined to represent the residuals between feature points in the monocular image and the corresponding points resulted from projecting the human model to the projection plane. A procedure is developed to recovery the whole body posture, and angle constraints are considered to obtain more natural human postures. Our model makes it feasible to reconstruct any possible human postures, and hence possible to generate highly realistic human animation. The experimental results from many monocular images are satisfactory.

Important Links:

Go Back