RETRACTED: A novel similarity metric for image filtering

Abstract

The image quality investigation is a significant and challenging task. One of the principal issues is the presence of noise that influences the nature of image. Denoising assume an essential part in image quality examination. The nature of the image is dissected in view of different execution measurements. This work centers on the measurement of similarity for the filtered image and the estimation is made on different normal and clinical pictures. Denoising is done by different image filtering draws closer, also the result is great for different phases of image handling like segmentation and classification approaches. An enhanced Discrete Wavelet Transform (EDWT) is utilized for image filtering and the proposed similarity measurements in view of sine measurement among the noise and the input image for image denoising and examining the nature of image. Alongside the current comparability measurements, different measurements like Peak Signal to Noise Ratio, Normalized Absolute Error, and Structural similarity Index Measure are considered for the work. The investigation is tried on various images like brain MRI scan images from the cameraman, Barbara, vegetables, remote sensing images, natural image as well as BRATS dataset,. The simulation outcome shows that the EDWT based filtering yields an improved outcome for the performance measurements.