Development of real-time monitoring system of automatically-extracted seafloor objects based on a degree of perceptual importance

2014 Fiscal Year Final Research Report

• Project/Area Number: 25820418
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Naval and maritime engineering
• Research Institution: The University of Tokyo
• Principal Investigator: YOSHIKI Sato 東京大学, 生産技術研究所, 特任研究員 (90635210)
• Project Period (FY): 2013-04-01 – 2015-03-31
• Keywords: 半余弦関数ウェーブレットネットワーク / 自律型海中ロボット / 誘目性に基づく物体抽出 / 空間周波数解析 / ハードウェア実装

Outline of Final Research Achievements

In general, observed image by AUVs (Autonomous Underwater Vehicles) can not be confirmed until the AUVs are recovered. It is because underwater communication devices are too low speed to send image data. In this study, real-time monitoring system of seafloor objects was developed based on the half-cosine wavelet network. The system enables automatical extraction of seafloor objects based on a degree of perceptual importance. The system adapts to the low speed and unstable acoustic communication by reducing the amount of data. The system can reduce the data size of image to less than 4%. Although obtained object image in real time are low resolution, distinguishable features are left.

Free Research Field

自律型海中ロボット