*2.3.1 Digital image correlation*

With the advent of digital cameras, time-lapse imagery has become popular since the beginning of the 2000s in glaciology, where it has been applied to survey polar ice flow [25–28] and mountain glaciers [6, 8, 29–33].

DIC is an image analysis technique that is applied to a pair of images to obtain spatially distributed maps of the two displacement components orthogonal to the line-of-sight (LOS). In classical DIC processing, a reference template out of the master image is searched for in an investigated larger area of the slave image. The cross-correlation (CC) is calculated for every possible template of the investigated area and the position of the maximum correlation coefficient corresponds to the displacement of the master template. Alternatively, the CC can be calculated in the Fourier domain according to the convolution theorem. Fourier CC is computationally efficient but it is more prone to outliers.

The main DIC advantages concern the low-cost hardware and its high portability even in harsh environments. Nevertheless, it suffers adverse meteorology and it strongly depends on the visibility conditions.
