Learning, tracking and recognition of 3D objects

摘要: In this contribution we describe steps towards the implementation of an active robot vision system. a sequence images taken by camera mounted on hand robot, detect, track, and estimate position orientation (pose) three-dimensional moving object. The extraction region interest is done automatically motion tracking step. For learning 3-D objects using two-dimensional views estimating object's pose, uniform statistical method presented which based expectation-maximization-algorithm (EM-algorithm). An explicit matching between features several not necessary. acquisition training required for process needs correlation image object its pose; performed robot. robot's parameters are determined hand/eye-calibration subsequent computation position. During estimation stage computed active, elastic contours (snakes). We introduce new approach online initializing snake first given sequence, show that snakes suited real time tracking. >