Translator Disclaimer
25 September 2002 Automated registration of polarimetric imagery using Fourier transform techniques
Author Affiliations +
Polarimetric imagery that is collected from time-sequential and multiple image format sensors all have potential for image misregistration. Since polarization is usually measured as small differences between radiometric measurements, it is highly sensitive to misregistration, especially at regions of high contrast. The general consensus in the polarization community is that image misregistration on the order of 1/10th of a pixel can introduce artifacts in polarization images. If the registration is not achieved and maintained to this resolution, the data must be registered in software. Typically, rotation and translation (horizontal and vertical) are the main transformations that need to be corrected. It is desirable to have a registration algorithm that determines rotations and translations to 1/10th of a pixel, does not require user intervention, takes minimal computation time, and is based on analytical (non-iterative), automated calculations. This paper details an analytical, automated registration algorithm that corrects for rotation and translations by using a Fourier transform technique. Examples of images registered with this algorithm, and estimates of residual misregistrations are presented. Typical processing times are also given.
© (2002) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Christopher M. Persons, David B. Chenault, Michael W. Jones, Kevin Deane Spradley, Michael G. Gulley, and Craig Alan Farlow "Automated registration of polarimetric imagery using Fourier transform techniques", Proc. SPIE 4819, Polarization Measurement, Analysis, and Applications V, (25 September 2002);

Back to Top