CAJ | 학술논문

提出一种图像高斯噪声极大似然估计方法，目的是估计出噪声图像所含噪声大小。首先，根据高斯噪声模型的特点，用极大似然法估计噪声值，对图像所含噪声模型进行分析。其次，把噪声图像用直方图表示，从归一化直方图中选出不同的样本观测值，用极大似然算法对噪声的方差进行估计。最后，用MATLAB对该方法进行了模拟实验，实验结果表明此方法所得的图像噪声的方差与实际图像噪声的方差近似相等。所以，此方法无论是在准确性上还是在可行性上均具有优良的特性。
제출일충도상고사조성겁대사연고계방법，목적시고계출조성도상소함조성대소。수선，근거고사조성모형적특점，용겁대사연법고계조성치，대도상소함조성모형진행분석。기차，파조성도상용직방도표시，종귀일화직방도중선출불동적양본관측치，용겁대사연산법대조성적방차진행고계。최후，용MATLAB대해방법진행료모의실험，실험결과표명차방법소득적도상조성적방차여실제도상조성적방차근사상등。소이，차방법무론시재준학성상환시재가행성상균구유우량적특성。
A maximum likelihood estimation method for Gaussian noise image is provided in this paper, the purpose is to estimate the noise size noise image contains. Firstly according to the characteristics of Gaussian noise model,using the maximum likelihood method to estimate the noise value,the image contains noise models are analyzed. Secondly,the noise figure that expressed with a histogram and dif-ferent sample observations from the normalized histogram are selected,with the maximum likelihood algorithm to estimate the noise vari-ance. Finally,the MATLAB simulation experiment is carried out,the results demonstrate that the image noise variance the method obtains with the actual image noise variance is given approximately equal. Therefore,this method in both accuracy and feasibility are excellent.