基于小波变换的图像去噪方法研究详解.docVIP

毕业设计（论文） 基于小波变换的图像去噪方法研究 院 别 专班级姓名 指导教师 2014年6月10 日 基于小波变换的图像去噪方法研究 摘要 Research on Image Denoising on Wavelet Transform Author： Tutor：Abstract Generally speaking, the images in our real life always contain noise. Therefore，for better subsequent processing, it is necessary to denoise the images.However, the traditional way of denoising the images is an obvious contradiction which aims at smoothing noise of images as well as retaining the details in the images. Thus, it has become a common goal of many researchers to find a way that can not only denoise images but also preserve the images details. Through research and practice，we can find wavelet transform can reduce the noise, and meanwhile retain edge information of the images well. So, we discusses the denoising algorithm based on wavelet transform in this test.In various denoising algorithms based on wavelet transform, this text primarily discusses wavelet threshold denoising and the wavelet transform modulus maxima, and test the two methods by simulation then analyze. By testing the threshold function by simulation, it can be found that eclectic function of soft and hard thresholding has better effect on images that are polluted by poisson noise, salt and pepper noise, gauss white noise and speckle noise, while semi-soft threshold seems a better choice for denoising the images which require to remove little noise and preserve more detail information. At the same time, through the experimental study we can also find wavelet transform modulus maxima is efficient to denoise different kinds of noises, especially to denoise the low SNR images. Nonetheless, since the mainstream algorithms are inefficient, wavelet transform modulus maxima in general cannot receive satisfactory results. Key Words: Image de-noising, Wavelet transform, Thres