Real-time salt and pepper noise removal from medical images using a modified weighted average filtering
Shilpi Gupta, Ramesh Kumar Sunkaria · 2017
With the elaboration of telemedicine communication, requirement of medical images have been increased rapidly. X-Ray, CT Scan, MRI, Ultrasound etc. are some imaging techniques which are used to obtain these medical images. During the transmission of medical images, noise becomes a dominant factor, which deteriorates the quality of an image. Salt and Pepper Noise (SPN) is one of the common noise which occur in acquisition or data transmission through any network or any medium. In this paper we anticipated a method named as Modified Weighted Average Filter to denoise four standard images such as brain, knee, mammogram and head. The method firstly detect pixels which contain noise by selecting the optimal value of the mask or kernal and then apply the filter which computes the weights according to the correlation present between the corrupted noisy pixels and normal noise free pixels. Peak signal to noise ratio (PSNR) will obtain the results of this method on the basis of numerical measurement and will be compared with the existing methods. Simulation results shows that proposed method significantly remove the high density SPN (up to 90%) in the medical images.