Profil Anggota
A Study on HFR-Video-Based Behavior Assessment for Livestock Trembling
This dissertation develops a non-contact multimodal sensing framework for cattle trembling monitoring by integrating high-frame-rate video, Digital Image Correlation, frequency analysis, thermal imaging, and 3D body-surface reconstruction. The research is conducted in three stages: two-dimensional trembling detection using HFR video, model-based 3D vibration–thermal fusion, and extension to non-rigid cattle body-surface monitoring in a barn environment. The proposed framework enables objective detection, localization, and representation of trembling-related motion together with thermal and geometric information. Experimental results show localized trembling response mainly around the abdominal region, with dominant frequency components in the 9.375–12.5 Hz band and visible body-surface temperature ranging approximately from 20 to 33°C. This dissertation does not aim to provide disease-specific diagnosis or complete 3D vibration-vector recovery. Instead, it establishes a quantitative, non-contact, and geometry-aware sensing foundation for future health-oriented livestock monitoring.
Riwayat Publikasi/Konferensi/Exchange/Internship atau sejenisnya 1
Model-Based Multi-Modal Volumetric Sensing via Fusion of Stereo HFR, ToF, and Thermal Cameras
Publikasi 1