Method, apparatus and program for image processing

Abstract

Image processing comprises obtaining (110) an image data set, the set being associated with parameter(s) representative of image capture characteristic(s) (e.g. exposure) for the set and comprising pixel intensity values representing image pixels having respective pixel locations in an image. For a given pixel intensity value in the set, an estimated noise value is determined (120) based on the parameter(s) associated with the set (e.g. using the square root of the parameter), and a representative intensity value derived from pixel intensity value(s) in the set (e.g. using neighbouring pixel values). The estimated noise value is associated (130) with the given pixel intensity value (e.g. by creating an index) and used to e.g. denoise the pixel intensity value, determine a compromised pixel intensity value (e.g. stuck, hot, dead or otherwise defective pixels), or provide the noise value and image to a computer vision system. The representative intensity value is intended to provide a measure of what the given pixel intensity value would be if there was no noise. The representative intensity value may comprise the given pixel intensity value.

Images