Vibration monitoring using sub-pixel edge localization

The main aim of our work presented in this article was to verify the possibility of a real-time monitoring of mechanical vibrations or shocks using edge detection from the images of the moving object recorded by a digital camera. Two sub-pixel edge detection methods designed for a 1-D images, grey level moment (GLM) edge detector and edge localization based on approximation of real image function with erf function (AEF) are investigated. The presented experimental results confirmed that the subpixel edge detection methods have the potential to track even small movements of the captured object, not exceeding two pixels. In addition, presented findings confirm that, under certain conditions, the blurring parameter of the edge model can be used to estimate the velocity of a moving object from a single frame.