RADIOMETRIC CCD CAMERA CALIBRATION AND NOISE ESTIMATION

Changes in measured image irradiance have many physical causes and are the primary cue for several visual processes, such as edge detection and shape from shading. Using physical models for charged-coupled devi ce (CCD) video cameras and material reflectance, we quantify the varia tion in digitized pixel values that is due to sensor noise and scene v ariation. This analysis forms the basis of algorithms for camera chara cterization and calibration and for scene description. Specifically, a lgorithms are developed for estimating the parameters of camera noise and for calibrating a camera to remove the effects of fixed pattern no nuniformity and spatial variation in dark current. While these techniq ues have many potential uses, we describe in particular how they can b e used to estimate a measure of scene variation. This measure is indep endent of image irradiance and can be used to identify a surface from a single sensor band over a range of situations. Experimental results confirm that the models presented in this paper are useful for modelin g the different sources of variation in real images obtained from vide o cameras.