Abstract

The correlation of two images or other spatial signals is a measure of their mutual similarity, hence useful for image interpretation, pattern recognition, etc. Simple in concept, but slow in performance is the method of direct correlation, where the two image transparencies are placed in contact and the total light throughtout is measured. This process is time consuming because of the need to repeat the measurement with different lateral shifts between the two transparencies. We have performed image correlations whereby a continuous range of shifts was automatically implemented. A transparency was illuminated with white light and followed by a dispersive element. Each wavelength then produced an image with a different lateral shift. After passing through a second transparency the spectral distribution represented the correlation of the two input functions.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call