Object recognition

Finding, tracking and analyzing objects is the core functionality in dotCube. For this purpose, dotCube provides a very wide range of functions and settings, allowing you to process large amounts of data in an extremely short time. Robust algorithms developed in collaboration with experimental physics laboratories provide a precise determination of the results.


Objects can be searched automatically in images, image sequences and movies with dotCube. Therefore, you either define a set of geometric properties yourself or make use of our search wizard: You frame an object and the wizard selects a geometric set of properties to find similar objects. Two complimentary algorithms are available: object search in respect to the local or global height differences between object and subsurface. Build-in noise reduction provides excellent results even from disturbed recordings, as they occur in everyday experimental life.


Similar objects within image sequences and movies are searched and identified by dotCube. Each object is given a unique label, a so-called ID, in order to be rediscovered and tracked any time. The object tracking is controlled by detailed settings which ask questions about the drifting behaviour:

Is the number of objects constant?
Are the objects moving?
Is the background moving?
Is there a reference object?
With the help of the object tracking, you are able to process large amounts of data very quickly.


Found objects can be characterized by more than 120 properties. The results are stored as meta data in each image and are available any time in the future. The object characteristics are both visual displayed as an extra layer within the image and as numbers in a result table which can be used for further data processing with spreadsheets and plot functions.