在信息时代数字图像处理已经成为人们获取复杂信息的重要手段，在一定程度上拓 展了人类的视觉。在众多领域得到了广泛应用，取得了巨大的社会效益和经济效益，因 此，图像处理技术的相关研究和应用无论从理论上还是从实践上都具有重要意义。 在数字图像处理领域，图像噪声的滤除一直是最重要、最基本的研究课题之一。由高斯噪声和椒盐噪声叠加而成的混合噪声是数字图像中存在的一种典型噪声。而传统方法对于这种类型噪声的处理效果往往是不尽如人意的，主要表现在滤除图像噪声的同时会对图像细节产生丢失。 绝大部分自然图像几乎同时含有椒盐噪声和高斯噪声，简单的使用传统的滤波算法不能获得理想的滤波效果。为了解决混有这两种噪声图像的滤波问题，分别针对以椒盐噪声为主的混合噪声图像 和高斯噪声为主的混合噪声图像，提出了双阈值滤波算法。 这种算法是在修正后的阿尔法均值滤波算法的基础上做了两方面的改进：首先，提出在图像邻域内为不同灰度值的像素点给出归一化的权值，用这些权值和其对应的灰度值共同决定滤波输出。其次，所设计的权值可以用修正因子来进行微调，来获得理想的滤波效果。实验证明，其处理效果优于传统滤波算法和修正后的阿尔法均值滤波算法。 本论文所提出的算法均在MATLAB上进行了仿真，并有相应的程序和截图，从而确保论文的真实性和算法的优越性。通过仿真结果可以发现，本文提出的算法在主观评价指标和客观评价指标上均符合要求。

Digital image processing has become an important method of aquiring complex information in an information age，and it has extended the vision of human beings in some extent．At present，digital image processing has been applied comprehensively in lots of fields and acquired enormous social and economic benefits. Therefore,the relative research and application of the digital image processing makes a significant sense for theory and practice． In the field of digital image processing, image denoising is one of the most important and basic research topic at all times. The mixed noise composed by Gaussian noise and Salt and pepper noise is a typical noise model in digital image. Traditional methods used to resolve this kind of noise are often less satisfactory than expect, which appears that some image details may be lost while doing image denoising． Most natural images containing gauss noise and salt and pepper noise at the same time, the use of simple traditional filtering algorithms can not get ideal filtering effect. In order to solve the problem of mixed noise of image filtering, respectively to give priority to with Gaussian noise and mixed noise image is given priority to with salt and pepper noise mixed noise of image, Proposed dual-threshold filtering algorithm These two kinds of algorithm is revised in alpha average filtering algorithm improvements made on the basis of two aspects: first, in the image in the neighborhood for different grey value of pixels are normalized weights, using these weights and their corresponding grey value to filter the output. Secondly, the weights can be designed to fine-tune with correction factor, to obtain the ideal filter effect. Experiments show that the treatment effect is superior to the traditional filtering algorithm and the adjusted alpha average filtering algorithm.In this thesis, experiments have been done in the PC of software MATLAB and experiment results indicate that the fingerprint verification algorithm is more effective, feasible, and it achieves the desired results.