Efficient Noise Removal Based on Non-Local Means Filter and Its Method Noise Wavelet Packet Thresholding

Image denoising includes operation of the image data to yield a visually high quality image and a major process in image processing, pattern recognition, and computer vision fields. The chief aim of image denoising is to re-establish the original image from a noisy image and help the other system (or human) to understand it well. Even though introduces many methods there only reduced visual quality, causing blurring and artifacts in the image. In this project, uses a new technique called wavelet packet transform and adaptive wavelet thresholding to denoise the image and improve visual quality. wavelets based denoising method consist of three steps namely, first to compute the wavelet packet transform, the next step to Remove noise from wavelet coefficients using wavelet packet thresholding and the third step is to reconstruct the enhanced image using inverse wavelet packet transform. Unlike standard wavelet-based methods Wavelet packet transform (WPT) used for image decomposition. The proposed method namely, Fast OWB extraction is a new adaptive thresholding function introduced to improve the denoising efficiency. Hence chooses an adaptive threshold value which is level and subband dependent based on analyzing the sub band coefficients. Estimation of dominant coefficients based on Maximum a posteriori (MAP) estimate to enhance or eliminate the wavelet coefficients. The resultant method yields better peak signal noise ratio with visual image quality measured by universal image quality index compared to standard denoising methods.