Scene interpretation by fusion of segment and region information

摘要

This paper describes a system that gives a simplified interpretation of real scenes through analysis of monocular images. Scenes are described by fusing information derived by the edge segments and by the regions extracted from the scene. The interpretation is given in terms of primitives such as surfaces and their spatial relationships. Real indoor and outdoor scenes relative to urban and rural environments have been used to validate the system. The main goal of the proposed architecture is to provide help for blind people's navigation, but the techniques developed are also suitable for robotics applications.