Automatic estimation of the noise model in fundus images

For several years, a lot of studies have been done on image analysis and image understanding. In medical imaging, the retinal images are usually corrupted by noise in its acquisition or transmission. There are various sources of noise inherent in the use of CCD's (Charge Coupled Device) and other external effects such as space radiation (cosmic rays) can have negative effects on the obtained data. In the context of retinal image denoising, most of algorithms assumes the noise is additive and independent of the RGB image data, and is also a Gaussian sample. However, the type and level of the noise generated by digital cameras are unknown if we don't have enough informations about the sensor and circuitry of a digital camera. Therefore, these approaches cannot effectively recover the ”true” signal (or its best approximation) from these noisy acquired observations. Thus, modeling noise in retinal images is an important and huge step before any processing task. The purpose of this paper is to explore CCD's and understand an algorithm for estimating the noise model in fundus image.