An Optimised Fuzzy Approach to Remove Mixed Noise from Images

Abstract

Mixed noises can be defined as a combination of different types of noises acting on a single carrier. There has been a mention of various mechanisms used to restore images corrupted with mixed noise in the past. This paper proposes a simple method based on fuzzy set theory and Bilateral Filter to remove mixed noises and compares it with previously mentioned techniques such as: Vector Median Filter(VMF), Vector Direction Filter (VDF), Fuzzy Peer Group Averaging (FPGA), Fuzzy Vector Median Filter (FVMF), Bilateral Filter (BF), Adaptive Bilateral Filter (ABF), Switching Bilateral Filter (SBF), Joint Bilateral Filter (JBF), and Trilateral Filter (TF) on the basis of performance metrics such as Peak Signal to Noise Ratio (PSNR), Mean Absolute Error (MAE), Mean Square Error (MSE) and Normalised Colour Difference (NCD). For the purpose of a detailed analysis, the performance of each method is evaluated by varying the image size and the noise density by implementing them in MATLAB-09. The mixed noise used in this paper is a combination of three noise i.e. poisson, impulse and Gaussian noise. The simulation and result shows that the proposed method provides better PSNR and hence better image quality than almost all the methods mentioned above.