Image Quality Assessment Parameters for Despeckling Filters

Images Quality assessment shows vital role in image processing applications and is an active field of research. In medical imaging field various imaging modalities i.e. Ultrasound (US), X-ray, Computer Tomography (CT) and Magnetic Resonance Imaging (MRI) etc. are used to diagnose the disease by the doctors. In US imaging modality coherent sources are used for imaging purpose. Due to coherent sources, speckle noise inserts in US images and causes loss of information and blurring which leads to misdiagnose of the patients by the doctors. To avoid this situation, various despeckling filters are used to remove the speckle noise and their performance is analysed on the basis of subjective and objective assessment parameters. This paper summarizes the various performance evaluation parameters i.e MSE, RMSE, SNR, PSNR, AD, SI, NK, MD, LMSE, NAE, IQI, SSIM, and BETA, their significance & relevance with the example of despeckling filters i.e. LEE Sigma, Detail Preserving Anisotropic Diffusion (DPAD), Fourier Butterworth Filter (FBF), ROFTV (Rudin, Osher and Fatem) and Hybridized Mean and Wiener Filter.