RUS  ENG
Full version
JOURNALS // Computer Optics // Archive

Computer Optics, 2021 Volume 45, Issue 2, Pages 267–276 (Mi co907)

This article is cited in 12 papers

IMAGE PROCESSING, PATTERN RECOGNITION

A method for measuring digital camera noise by automatic segmentation of a striped target

N. N. Evtikhiev, A. V. Kozlov, V. V. Krasnov, V. G. Rodin, R. S. Starikov, P. A. Cheremkhin

National Research Nuclear University MEPhI (Moscow Engineering Physics Institute), 115409, Moscow, Russia, Kashirskoye Shosse 31

Abstract: Currently, cameras are widely used in scientific, industrial and amateur tasks. Thus, one needs to be able to quickly evaluate characteristics and capabilities of a particular camera. A method for measuring noise components of the camera photosensor is proposed. It allows one to estimate shot noise, dark temporal noise, photo response non-uniformity and dark signal non-uniformity. For noise measurement, just two images of the same scene need to be registered. The scene consists of several stripes (quasihomogeneous regions). Then the images are processed by automatic signal segmentation. The performance and accuracy of the proposed method are higher than or equal to other fast methods. The experimental results obtained are similar to those derived using a time-consuming standard method within a measurement error.

Keywords: digital cameras, sensors, digital image processing, image segmentation, shot noise, photo response non-uniformity.

Received: 25.09.2020
Accepted: 16.12.2020

DOI: 10.18287/2412-6179-CO-815



© Steklov Math. Inst. of RAS, 2025