京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第21910号 / 情博第693号 / 新制||情||119(附属図書館) / 京都大学大学院情報学研究科知能情報学専攻 / (主査)講師 延原 章平, 教授 西野 恒, 准教授 飯山 将晃 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

We describe a framework for global shipping container monitoring using machine learning with multi-sensor hubs and infrared catadioptric imaging. A wireless mesh radio satellite tag architecture provides connectivity anywhere in the world which is a significant improvement to legacy methods. We discuss the design and testing of a low-cost long-wave infrared catadioptric imaging device and multi-sensor hub combination as an intelligent edge computing system that, when equipped with physics-based machine learning algorithms, can interpret the scene inside a shipping container to make efficient use of expensive communications bandwidth. The histogram of oriented gradients and T-channel (HOG+) feature as introduced for human detection on low-resolution infrared catadioptric images is shown to be effective for various mirror shapes designed to give wide volume coverage with controlled distortion. Initial results for through-metal communication with ultrasonic guided waves show promise using the Dynamic Wavelet Fingerprint Technique (DWFT) to identify Lamb waves in a complicated ultrasonic signal.