Output-only computer vision based damage detection using phase-based optical flow and unscented Kalman filters

详细信息    查看全文

• 作者：Y.-J. Cha^a ; ¹ ; ^{young.cha@umanitoba.ca} ; J.G. Chen^b ; O. Bü ; yü ; kö ; ztü ; rk^c
• 关键词：Motion magnification ; Phase-based optical flow ; Video ; Computer vision ; Non-contact measurement ; System identification ; Unscented Kalman filters
• 刊名：Engineering Structures
• 出版年：2017
• 出版时间：1 February 2017
• 年：2017
• 卷：132
• 期：Complete
• 页码：300-313
• 全文大小：2701 K
• 卷排序：132

文摘

New innovative non-contract based damage detection method is developed. Used advanced computer-vision techniques to measure absolute displacement responses. Unscented Kalman filter was used to predict structural stiffness and damping. Various connection damage and member damage scenarios have been generated for experiments. Carried out extensive experimental studies to validate proposed method.