计算机技术与发展2017，Vol.27Issue(3)：81-84,4.DOI:10.3969/j.issn.1673-629X.2017.03.017
小波变换的阈值图像去噪算法改进
Improvement of Threshold Image Denoising Algorithm with Wavelet Transform
摘要
Abstract
There are many problems for wavelet threshold denoising method proposed by D. L. Donoho and constructed wavelet threshold function mentioned in this paper in image denoising through analysis and research. In order to improve these problems,a new improved threshold function has been presented which integrates the advantages of classical wavelet threshold function and other improved methods. This function is not only continuous at a specific threshold, but also involves parameters. Thus, constant deviation between threshold wavelet coefficients and original wavelet ones can be adjusted by regulating the parameters. Meanwhile,it is of differentiability convenient for calculations. In order to highlight the advantages of the threshold function constructed,Mean Square Errors ( MES) and Peak Signal to Noise Ratios ( PSNR) of several wavelet denoising methods are compared with proposed method in simulation experiment. Results of the experiment show that the images denoised with new threshold function are better than those with traditional soft and hard threshold func-tions and existent threshold functions,either in visual effect or performance parameters such as MES and/or PSNR.关键词
小波变换/阈值函数/图像去噪/均方差/峰值信噪比Key words
wavelet transform/threshold function/image denoising/MSE/PSNR分类
信息技术与安全科学引用本文复制引用
张小燕,吐尔洪江·阿布都克力木..小波变换的阈值图像去噪算法改进[J].计算机技术与发展,2017,27(3):81-84,4.基金项目
国家自然科学基金资助项目(11261061,61362039,10661010) （11261061,61362039,10661010）
新疆维吾尔自治区自然科学基金资助项目(200721104) （200721104）
