This work discusses a method for image based displacement identification based on spatial phase information. Some already known techniques and their properties are presented. Continuous wavelet transform is described, the code for its computation is written in the Python programming language, simplified with the use of convolution and further advanced for sub-pixel accuracy of the displacement identification. This function, together with the correlation method is then used for displacement calculation; firstly, for a one dimensional signal and afterwards on a series of images, both synthetic and filmed with a high-speed camera. The influence of noise was researched, possible difficulties and further progress were described.
|
