Position encoding and localization with environmental patterns

In this paper we discuss the design and development of an improved CLUSPI method for augmented computer vision and positioning of autonomous agents in indoor settings. The method employs environmental patterns posted on walls, ceilings, floors, and other surrounding surfaces that are accessible for digital imaging. Such patterns are blended into the environment as decorative elements where the encoding and decoding is based on orientation and clustering of artistic figures. As part of this work a specialized client-server system for multi-platform experiments with various environmental codes and imaging devices have been implemented. Conducted experiments indicate robust and reliable code extraction with very high recognition rates in most practical setups.