Real-Time Video Surveillance Based Structural Health Monitoring of Civil Structures Using Artificial Neural Network

Abstract

Modern world’s incessantly increasing outdoor traffic load has eventually led to structural health concern and continuous health monitoring of large scale civil structures such as bridges, roads, highways, etc. In this paper, we propose a computer vision based non-destructive structural health monitoring (SHM) method using high speed camera system combined with the brilliance of artificial intelligence. A number of appreciable SHM techniques had been reported that utilizes wired or wireless smart sensors, but the use of nondestructive techniques, such as, digital high speed imaging were rarely employed for detection of dynamic vibrations of civil structures. In the current research, we have developed a high speed video imaging based structural health monitoring system that utilizes blob detection based motion tracking algorithm. It provides factual information regarding localization and displacement of the target object or an existing feature in the civil structure. The modal parameters were subsequently extracted to analyze the level of severity of structural damage within the civil structures. Also, an artificial neural network is trained to infer the qualitative characteristics of structural vibrations based on vibration intensity and the network inferences can be correlated with the conditions of the structure. The efficacy of our vision system in remote measurement of dynamic displacements was demonstrated through a shaking table and a slip desk experiment. The experimental results demonstrate real-time output with satisfactory performance.