THREE-DIMENSIONAL OBJECT RECOGNIZING APPARATUS BASED ON IMAGE DATA

摘要

PURPOSE: To enable the generation of a three-dimensional image of an object as well by recognizing an object after the estimation of a three-dimensional motion and a position attitude of the object. CONSTITUTION: A model data base 7 has three-dimensional wire frame information of an object stored and an moving object extracting section 4 extracts an area of a moving object by a block matching method from an image source with a CCD camera 2. A space quantizing section 10 estimates a three- dimensional position attitude of a moving object performing a comparative computation of a two-dimensional image of the area of the moving object and an estimated image generated by a texture mapping. A texture mapping section 9 executes a texture mapping for a wire frame of the object with the wire frame information in an object memory 8 as reference and using position attitude information to be outputted from the space quantizing section 10 and an image estimating section 11 generates a virtual real image of the object based on such a texture mapping.