A Study on Effective Repetition of Bilateral Filter for Medical Images

Abstract

Recently, a camera has the function of gain bright- ness that makes a bright photo from the dark photo. The function is realized by amplifying an attenuated signal from a photosensor in a digital camera. But it amplifies noise with signal. The noise makes salt-pepper-noise on a monochrome photo; Gaussian noise on a color picture. In the previous study, the bilateral filter was used for denoise on a picture. The bilateral filter is a nonlinear filter that is able to denoise and save important corner, edge, and contours. The bilateral filter needs two or three parameters that are different to do best for each input pictures. The previous study searches the optimum parameter, but the method needs to be run the bilateral filter for each suggested parameter and the execution time was long. The bilateral filter has a character is able to remove noise effectively by filtering several times than search optimum parameters. In this article, our purpose is considered of the effective number of repetitions of Bilateral Filter. We estimate how many repeat filtering by using SSIM and PSNR for image quality measuring.