Towards image-based marine habitat classification

It is now fairly routine to quasi-automatically generate acoustic bathymetry and optical mosaics from properly instrumented Autonomous Underwater Vehicles (AUVs). However, further analysis and interpretation of gathered data is needed to address tasks such as habitat characterization and monitoring. This analysis stage is performed by human experts which limits the amount and speed of data processing. While it is unlikely that machines will match humans at fine-scale classification, machines can now perform preliminary, coarser classification to provide timely and relevant feedback to assist human decisions and enable adaptive AUV behavior. This paper presents a preliminary investigation into using a state-of-art object recognition system to classify marine habitat imagery based on labeled examples. We show that performance for such approaches can suffer with typical underwater imagery and present some of the causes for this. We propose modifications that make such a system suitable for automated coarse habitat classification and discuss experiences and results with three applications. The first corresponds to towed imagery from Ningaloo and Scott Reef, Western Australia. The second corresponds to AUV imagery near Hydrographers passage, Queensland. The third application demonstrates adaptive surveying using the output of the modified classification system.