Object Recognition Group
It is difficult to recognize some objects in the image for automatically. However, information of three dimension shapes and the objects are important to behave to them. Then, it is necessary to recognize automatically three dimension design of various objects and the environments in "Friendly Informative Cyber Space". For this purpose, we research on the Object Recognition by combined information on analysis of line drawing image, Image Segmentation, Stereovision and so on.
Core Techniques: Object Recognition Method Based on 3D Model, Contour state of object・Graph, CG（Shade, VRML),Machine Learning
Research example 1: Automatic scene analysis for soccer movie
Research example 2: Security Key by using hand shape
Research example 3: Person certification by using fingertip sign