A New Performance Metric Algorithm in Image Filtering

Denoising play a vital role in image quality analysis and it is regarded as a challenging task. The quality of the image is analysed based on various performance metrics. This paper focuses on the quality metric for image filtered output and the testing is made on various general and medical images. Denoising is performed by various image filtering approaches and the outcome is good for other stages of image processing such as segmentation and image classification. An enhanced Discrete Wavelet Transform (Enhanced DWT) and a novel performance metrics based on logarithmic estimation are proposed for image denoising and analysing the quality of image. Along with the proposed similarity metrics other metrics like PSNR, NAE, and SSIM are consider for the work. The performance analysis is tested on different images like biomedical images, Barbara, cameraman, vegetables, and remote sensing images as well as a natural image. The experimental result shows that the proposed enhanced DWT yields a better result for all the similarity metrics.