Showing papers on "Dark-frame subtraction published in 1981"

Techniques For Speckle Noise Removal

Abstract: In this paper, several techniques to reduce speckle noise (more generally signal independent multiplicative noise) in images are studied. The techniques include gray scale modification, frame averaging, low-pass filtering in the intensity and density domains, and application of the short space spectral subtraction image restoration technique in the density domain. Some discussions on the theoretical basis of the techniques studied are given and their performances are illustrated by way of examples.

Noise reduction in digital fluoroscopy systems

Abstract: Apparatus for reducing the effect of x-ray statistical noise and electronic noise in a fluorographic system that displays an x-ray image on a television screen. Analog video signals based on the x-ray image are amplified logarithmically and digitized to yield live pixel signals. Processed pixel signals are averaged in a full image store or memory. Motion is detected by subtracting the stored pixels from the live pixels on a pixel-by-pixel basis in an ALU. The difference resulting from subtraction is used as part of an address to a look-up table (LUT) which contains values equivalent to the difference signals times a noise reduction multiplicative factor, K. The other part of the addresses is the live pixel value. There are several replications of the look-up tables each relating to a particular brightness level range. The one selected is determined by the live signal part of the address which relates to brightness. The K times the difference signals in the ranges are chosen so the amount of noise reduction varies with brightness level as desired for logarithmic signals. The pixels processed as explained above are added in-phase with the stored and averaged pixels and returned to the corresponding full image memory locations.

Closed form bandlimited image extrapolation

Abstract: Digital implementations of a closed form 2-D bandlimited image extrapolation algorithm are presented for a number of elementary target images. The resulting figures of merit empirically verify the assumption that extrapolation is better near to where the image is known. Results of extrapolating a truncated image perturbed by white Gaussian noise suggest a required SNR of the order of 1010.

Signal-To-Noise Measurements On Channel Image Intensifier

Abstract: Second generation image intensifiers are generally used with cathode illuminations so low that the system performance is significantly degraded by quantum noise. The noise arises from the inherent quantum nature of the detected radiation and the statistical variation in the gain of the intensifier. A signal to noise measurement made under specified conditions provides a quantifiable assessment of this characteristic for any image intensifier. The conditions that are now generally used in specifications are given and a method of measurement described. Also methods of determining the individual contributions to the overall signal to noise ratio are described.