Image calibration in the case of integral translation and linear elongation

Abstract

A technique for image alignment with global translation and linear stretch determines translation parameters for three corresponding linearly displaced blocks in a reference image and a corresponding distorted test image. From the differences between the translation parameters for the three blocks the presence of stretch is detected and, if detected, a stretch factor is estimated. The estimated stretch factor is used as a starting point to stretch the reference image to overlap the distorted test image as a refinement process. The resulting refined stretch factor is then used in a reverse stretch process to shrink the distorted test image, and the distorted test image is then aligned with the reference image to obtain picture quality metrics. <IMAGE> <IMAGE>

Description

technical field [0001] The present invention relates to image calibration and, more particularly, to calibration by detecting and correcting extensions of images in the event of overall translation and linear extension. Background technique [0002] Accurate spatial alignment of digital images is a fundamental focus in many applications of image processing. The book "Computer Image Processing and Recognition" by Ernest Hall, Academic Press, 1979, introduces the basic correlation method of template matching that can be used for image calibration. Kuglin et al. published a paper entitled "Phase Correlation Image Calibration Method" in the scientific report collection of IEEE (Institute of Electrical and Electronics Engineers) 1975 International Conference on Cybernetics and Society in September 1975, pp. 163-165. The basic correlator scheme of information. These methods work well for global translations such as Figure 1A and 1B The kind that can be very precise. But for t...

AI Technical Summary

Problems solved by technology

But the size of the subregions must be quite large to avoid confusion problems, so

Images