The Digital Image Correlation (DIC) is an optical-numerical remote sensing technique that allows the measurement of "full-field" displacement and deformation on the surface of objects, through the correlation of co-registered images, collected at different time intervals by high resolution digital cameras.
The operational principle is based on the analysis of color changes in single pixels (or pixel groups), thus allowing the measurement of surface deformations of the ground (landslides, volcanoes, subsidence etc.) or structures, without any additional installation of contact sensors or reflectors.
Experimental applications have confirmed, through the integration with other remote sensing techniques, the high reliability of the techniques in both natural hazards and structural field.
Structure from Motion (SfM) is a computational technique that allows the reconstruction of 3D digital models, through the use of optical images with an high percentage of overlapping.
Simultaneously, all the information extracted from the images is determined by redundant iterative procedures based on a database which is automatically present in the metadata file.
For this reason, this approach is best suited for image sets characterized by a high percentage of overlapping and captured by as many different observation points as possible around the object (as also suggested by the designation of the same technique, Structure from Motion).
For Change Detection technique is meant the process of identifying, describing and quantifying variations of an object, phenomenon or area, occurring in a particular time interval.
This is possible by analyzing differences and variations between optical images, acquired with the same scene, at different times or conditions.
The fundamental assumption found in Change Detection analysis, carried out with remotely sensed data, is that the surface or ground changes are identified by analyzing the variations in radiometric behavior.
Thanks to the technique of Change Detection and the use of high resolution data, you can identify all the small variations, otherwise unrecognizable using traditional instrumentation.