• Pattern Match

    Correlation algorithms, sub-pixel accuracy, performance optimisation, avoiding false positives are the challenges to be solved. The picture shows part of the test GUI for the pattern matcher with pan and zoom facilities for the different images, for visual inspection of the result.

    By clicking into one of the pictures, you get a short description of the corresponding item.

    The following main technologies are used:

      • Phase Correlation: Algorithm supported by the Fast Fourier Transform (FFT) of the OpenCV library.
      • Sub-Pixel Accuracy: Achievable with proper interpolation algorithms.
      • Optimisation: Code Reduction supported by systematic time measurements.
      • Avoidance of False Positives: Effective, own-developed scoring algorithms