Abstract
The recognition of 3-dimensional space information is essential if robots are to help people in daily life. We propose a method that gets space information between camera and objects from quantity of movement in video camera and movement of characteristic in video camera images.