基于小波边缘检测的图像去噪方法研究-控制理论与控制工程专业论文.docxVIP

郑州大聿工学硕士学位论文(3)针对传统去噪方法不能保留边缘特征和全局阈值法。过扼杀”小波系数的 郑州大聿工学硕士学位论文 (3)针对传统去噪方法不能保留边缘特征和全局阈值法。过扼杀”小波系数的 缺点，提出了基于边缘检测和MAP估计的小波阈值去噪方法，该方法在去 噪之前，先通过小波边缘检测方法确定图像边缘位置，这些边缘点将不受阈 值去嗓的影响。同时闺值函数的选择也很重要，本文选用软阈值并进行仿真 比较。 关键词：阈值去噪小波变换边缘检测自适应样条小波 Ⅱ 基于小渡边缘检测的图像去噪方法研究Abstract 基于小渡边缘检测的图像去噪方法研究 Abstract The real-lifeimages∽always corruptedbynoises．Fortheneed offurtherimages mlalysis and communication,the noises need to be reduced in image per-processing． Recently,with the improvement ofwavelet theory,wavelet analysis has p∞etrated into many fields．Wavelet trlmsform has the etmraeteristic ofmathematics micaoscopo, thus it call not only wipe offnoises but also retllill the original image details．Traditional denoising methodsⅢfilter noise,砒the cost of the image details being m口y This palⅫpresents a lw image denoising method based Oil edge detection and MAP estimation in order to reHlOVe the noises while retaining as many the important signal features∞possible． Traditional methods of cdge detection are base On one-order derivatives maxmmm,Or two-order derivative’s zeao-erossing．This kind of edge definition is very sensitive to noises．So edge detection should be eanied out in large scale,by smoothing the images．One ofthe shortcomings ofodge dete．．elion in large scale is that it’s difficult to locate edgc precisely,which will Imke mistakes in pattem recognition in edgc features．With multi-scale clmzacterization,warder analysis was widely used to multi-scale edge detection．We make good use of the properties of multi—scale cdge information to locate image甜gc Wim B-$pline wavelet．This cm obtain both better image cdge and higbelf accuracy The detected edgc coettleients will be protected from dcnoising．With sub-band laigla-frequeney components,global thresholding cm throw off key folmation of image, while local thresholding under the rule of Bayesian maximum probability a posteriori(MAP)with locality is calculated by local components，which provids better r,耐ormance．The denoising algorithm ofthis pap