A NOVEL DENOISING TECHNIQUE FOR MIXED NOISE REMOVAL FROM GRAYSCALE AND COLOR IMAGES

摘要

The real-time images acquired from cameras, CCTV, medical image scanners like Magnetic Resonance Imaging (MRI), Computerized Tomography (CT), Ultrasound (US) and X-ray etc., are often corrupted by noise. This noise may be a mixture of two or more noise types. In recent years, researchers concentrate on developing a denoising filter to suppress the mixed noises to improve the quality of the image. A novel algorithm that uses absolute difference, mean and median for the removal of mixed noise in image has been proposed in this article. The proposed filter is tested with the images induced by two types of noise mixed (Salt and Pepper and Gaussian noise) and three types of noise mixed (Gaussian, Salt and Pepper and Speckle noise) images. The performance of the proposed algorithm is compared with existing Fuzzy Based Filter (FBF), and Median Weiner Bilateral Filter (MWBF) algorithms. The test images used in this research work are Lena image, Iris eye images and medical images in grayscale Joint Photographic Experts Group (JPEG) format and also with the color images in four different image formats with mixed noise level ranging from 0.01 to 0.10. The experimented results show that the proposed algorithm yields better performance than the algorithms mentioned above. Peak Signal to Noise Ratio (PSNR) and Mean Square Error (MSE) are the metrics used in this comparative analysis.