Motion Magnification of videos for the diagnostics of buildings' state of damage

The assessment of the state of damage of modern or ancient buildings is extremely important in the engineering practice. Today, the focus is on the use of non-destructive methods, among which video-based diagnostics may play a major role because of many advantages with respect to contact and traditional methods, especially in the application to Cultural Heritage assets. An emergent methodology within this field is the motion magnification (MM) of recorded video footages. MM technique is based on algorithms able to magnify the movements of objects in a video by amplifying the acquired pixels signals, while keeping the objects topology. As a result, tiny movements of objects present in the video, invisible to the naked eye, are magnified becoming clearly visible. Besides, they can be processed for a quantitative analysis of objects dynamics. Importantly, the magnification can be operated in a selected range of frequencies. In addition to traditional quantitative methods (e.g. in the frequency domain) innovative methods can be applied to magnified videos, along to simple qualitative visual methods. Here we intend to show some of them.