A novel radiometrically accurate scene-based nonuniformity correction (NUC) algorithm is described. The technique combines absolute calibration with a recently reported algebraic scene-based NUC algorithm. The technique is based on the following principle: First, detectors that are along the perimeter of the focal-plane array are absolutely calibrated; then the calibration is transported to the remaining uncalibrated interior detectors through the application of the algebraic scene-based algorithm, which utilizes pairs of image frames exhibiting arbitrary global motion. The key advantage of this technique is that it can obtain radiometric accuracy during NUC without disrupting camera operation. Accurate estimates of the bias nonuniformity can be achieved with relatively few frames, which can be fewer than ten frame pairs. Advantages of this technique are discussed, and a thorough performance analysis is presented with use of simulated and real infrared imagery.
© 2003 Optical Society of America
(100.2000) Image processing : Digital image processing
(100.2550) Image processing : Focal-plane-array image processors
(100.3020) Image processing : Image reconstruction-restoration
(110.3080) Imaging systems : Infrared imaging
(110.4280) Imaging systems : Noise in imaging systems
Bradley M. Ratliff, Majeed M. Hayat, and J. Scott Tyo, "Radiometrically accurate scene-based nonuniformity correction for array sensors," J. Opt. Soc. Am. A 20, 1890-1899 (2003)