“Point (Marker) Tracking” is a measurement technique that detects, correlates, and “tracks” discrete visual references, known as “fiducials”, on a surface such as: coded targets, generic markers, or alternatively elements within a continuous pattern. The method is used primarily for motion analysis for the retrieval of kinematic quantities of movement, and also for vibration analysis in operation, deflection, and shape (ODS) investigations.
Absolute 3D-position and relative measurements for displacement (which is a prerequisite for vibration analysis), velocity, acceleration, and angular rotation (for motion analysis) can all be performed using point (marker) tracking.
The basic concept works on the principle of selecting (or in some cases, automatically detecting) a fiducial object in the reference image, which is then identified and correlated, temporally, across measurement image states. Through importing a calibration model of the projection parameters from multiple cameras into the correlation process, the 3D information of the tracked marker can be spatially retrieved.
Unique to Istra4D is that users can perform both; point (marker) tracking for discrete kinematic quantities, but also, full-field contour (shape), displacement, and strain measurement. This advantage is harnessed through the employment of a shape model matching algorithm for discrete marker tracking, and Digital Image Correlation (DIC) for strain measurement. Not to mention, that users can track both (36-different) coded targets, and also (up to 100) speckled facets/subsets per measurement.
Figure 2 – Coded target markers
With target tracking functionality, coded markers can be automatically detected, quickly tracked, and accurately measured. Using the shape matching algorithm, Istra4D is actually “trained (in advanced) to know what to look for”. This means that the markers can be both distinguished from one another and found through automatic recognition. Shape matching is also more accurate (for point marker tracking) than using speckled facets for DIC, since the design is both „embedded” in the software database and more absolute in detection via the single crosshair contrast variation.
Read more on Digital Image Correlation for information
