Integration of Chemical and Visual Sensors for Identifying an Odor Source in Near Shore Ocean Conditions

This paper develops new multiple sensor-based algorithms for identifying a chemical odor source in a near-shore and ocean environment via an autonomous underwater vehicle (AUV). Those algorithms implement two modules: source declaration and source verification, which are embedded in a subsumption architecture for chemical plume tracing (CPT). The source declaration module is based on chemical events detected by a chemical sensor, in combination with measured vehicle locations and fluid flow directions, while the source verification module uses a fuzzy color segmentation algorithm to process an image taken when the odor source is declared.