Fish escapes can be reduced to a minimum with the Autonomous Underwater Vehicle (AUV), which has been developed by the Hellenic Centre for Marine Research (HCMR). The small-sized device will be equipped with a high definition mini camera and is used constantly throughout the fish-cage inspection. The use of the HD camera helps detect net holes early enough to avoid fish escapes. The AUV will be connected to an IP network and upon any positive detection will be able to appropriately alert the designated personnel through e-mail, thereby triggering remedial actions. The AUV’s autopilot system has been integrated with computer vision aided navigation and guidance software to allow real-time correction of any accumulated errors in AUV positioning while inside the cage. The navigation scheme is based on an optical recognition system using photogrammetry fundamentals.