Core Image Denoising Based On Wavelet Transform And Median Filter

The core image denoising is the core image processing of follow-up key. Core image have some similarities with the general sense of the digital image processing, provide a further application of image processing, but the core image itself particularity and complexity puts forward the new needs to solve the research topic.Study of the current commonly used method of image noise filtering core-wavelet transform,the paper introduces the wavelet transform, median filtering method of combining containing a mixture of both Gaussian and impulse noise core image denoising. It can effectively remove impulse noiseand maintain the good image edge information by using the median filtering de-noising, and it can effectively remove the core image of mixed gaussian noise, and can keep the detail of image by using wavelet threshold denoising. This method for low SNR material quality improvement is very meaningful.This paper analyzes and studies the removal of core images mixed with some Gaussian noise and impulse noise approach, the first to use matlab software, the original image with different intensity impulse noise and Gaussian noise, and then two-dimensional images with noise median filter, according to wavelet decomposition algorithm will be used Sym median filtering the image after the wavelet decomposition of three-tier, and then select the appropriate threshold for the image de-noising, reconstruction constitute the final inverse transform de-noising images, through objective evaluation criteria ---Peak Signal to Noise Ratio to determine the quality of the reconstructed image.