Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device

The Algorithm Research of Image Denosing Based on Median Filter and Wavelet Transform

Author WangXiangJu
Tutor LiaoShaoJun
School Xi'an University of Science and Technology
Course Applied Computer Technology
Keywords Image Denoising Median filter Wavelet transform
CLC TP391.41
Type Master's thesis
Year 2008
Downloads 958
Quotes 19
Download Dissertation

Image signal in the generation, transmission and recording process , often by a variety of noise interference , which seriously affect the visual effect of the image , so using appropriate methods to reduce noise ( image noise filtering ) , is a very important the work . In this paper, image de-noising method of the wavelet filtering technology and median filtering technology . First, a comprehensive overview of the status and common method of image denoising . Wavelet threshold filtering method for wavelet filtering technology has been widely used due to the simple calculation . Wavelet threshold filtering technology threshold function and threshold selection conducted in-depth research , this paper, an improved threshold function and threshold simulation test results show that this improved method has better denoising effect . Secondly, the study of the conventional median filter , the median filter is improved , minimax median filter , as well as better than the effect of the adaptive median filter algorithm , in Comparative analysis of their advantages and disadvantages in this paper a median filtering algorithm based on impulse noise detection and study its main properties . Simulation test shows this method has better denoising effect and detail preservation performance . Then on the image filtering mixed with a variety of noise filtering method . Median filter and wavelet filtering in the removal of the advantages of impulse noise and Gaussian noise , given an image mixed noise de-noising method . Median filtering based on impulse noise detection method based on wavelet characteristics of impulse noise and Gaussian noise , impulse noise detected and filtering , and Gaussian noise pollution pixels using wavelet threshold filtering method to remove Gaussian noise . The simulation test proved the practicality and effectiveness of this method has better effect than a single filter . Finally, at the end of this thesis work to do a detailed summary and image de-noising method further research directions are put forward .

Related Dissertations
More Dissertations