Object tracking by comparing generated CG images

3-D object motion tracking method without feature extraction is necessary for monitoring human action in an usual civil-life scene. To realize a fast and robust object tracking method, we propose a model-based method using intensity images. At first, the 3-D shape and reflectance of the object are measured by using a rangefinder constructed with a camera and projector. Then, several CG images with varied object pose and position are generated and then compared to the input image. The pose and position of the object are estimated by minimizing the residual of the CG and input images. We made a special CG generator which is a precise simulator of the real camera to generate a CG image identical to the input image. We confirmed the ability of our method and achieved a 3DOFs real-time object tracking system in 4.5frame/sec speed.