河川流計測における画像解析技術の実用化
>100 Views
October 19, 24
スライド概要
Title in English: Development & Practical Applications of Image-Based Techniques for River Velocity Measurements
It is presented in Mar.2013, as master's course completion report meeting at the Kobe University.
土木技術者として社会インフラの維持管理業務に携わっておりました。今は第二のキャリアとしてAI・機械学習技術者としてメーカーにて仕事をしています。 学生時代、「動画像による洪水流の表面流速計測技術」について研究開発しておりました。
関連スライド
各ページのテキスト
河川流計測における画像解析技術の実用化 Development & Practical Applications of Image-Based Techniques for River Velocity Measurements Masato HONDA Research Supervisor I.Fujita Department of Architecture and Civil Engineering Graduate School of Science and Technology Kobe University
Background
Problem
Background
Camera Calibration object points
Camera Calibration
Camera Calibration (making geo-rectified image with ease) In the mountain, setting object points is a difficult situation.
Camera Calibration
Float-PTV
Float-PTV
Float-PTV
STIV Search line
STIV 22.0m -1.94 m/sec ① Set templates on STI. Next,calculate gradient tensor of intensity in each templates ② Histofram of gradient on STI ③ Calculate averaged gradient φ on STI by peak of histogram
STIV FFT STI -1.91 m/sec FFT STI with equalization -1.93 m/sec
STIV
STIV ① ⑩
How To Use NEW STIV software
How To Use NEW STIV software
How To Use NEW STIV software
How To Use NEW STIV software
Conclusion › This study deals with the developments of Image based Velocimetry measurements. › For this purpose, we firstly developed camera calibration and PTV software chasing floats. › Next we established calibration software assuming use in the mountain where surveyor cannot set object points. › Finally, we developed new STIV software. This soft is useful and accurate.