We organized the development of the project in five research lines. In the first one, an Underwater Autonomous Vehicle (AUV) will be built according to the special requirements previously described. The work will not consist in building a new robot from scratch, but in upgrading the robots that the University of Girona has to reach the requirements of the proposed project. The AUV will be able to navigate in an autonomous way while acquiring images of the ocean floor from a "visual" altitude, using visual sensors, like the trinocular camera system develop under the FOTOGEO project (Garcia et al. 2006b), but also acoustic sensors, like the sonar camera Echoscope. Using various sensors the quality of the final result data can significantly be improved applying sensor fusion techniques.
To achieve the goal of the project the second research topic provides the robot with techniques to process visual and georeference data online (during surveys and autonomously) which may be used for robot navigation (e.g. closing gaps in mosaics, planning tracks, etc.) and easy inspection of collected data at the end of a given survey. This data is further refined and enhanced applying sophisticated global alignment techniques and also sensor fusion approaches that require development.
Change detection completes the project goal by providing techniques for long term monitoring of interesting scientific areas by repeated mosaic acquisition and comparison of these mosaics, and providing an operational vehicle for the scientific community. Appropriate comparison algorithms will be developed to enable easy visual inspection for biological and geological purposes.