A Remote Mycological Assistant

Inspired in the active field of content based image retrieval (CBIR), we propose the design of a remote system for assisting mycological amateurs to categorize the mushroom samples they find in the nature. The query to a remote database will be performed through a mobile phone with incorporated digital camera. The system is a blend of CBIR and rule based expert system. In its current state of development we have implemented a shape based CBIR system complemented by a rule based expert system that achieves a high success rate over a in-house collected image database. The paper also describes the global system architecture.