毕业设计（论文）--基于图像识别的滤波器设计.doc

武汉科技学院毕业论文 论文题目：基于图像识别的滤波器设计 学 院 机电工程学院 专 业 机械设计制造及其自动化 班 级 机设057班 学生姓名 廖明健 学 号 0502051308 指导老师 朱里 2009年5月25日 基于图像识别的滤波器设计 摘 要 本文的研究课题是“基于图像识别的滤波器设计”。图像是人类获取和交换信息的主要来源因此，图像处理的应用领域必然涉及到人类生活和工作的方方面面。图像受获取和存储、处理及各种干扰的影响，显示时画面上会出现噪声。为了减少噪声带来的负面影响，尽可能地还原干净真实的画面，就需要用到降噪滤波器对图像数据进行处理。常见的降噪方法有在空间域进行的，也有将图像数据经过傅里叶等变换以后转到频域中进行的Abstract Research topic of this article is based on image recognition of the filter design. Image acquisition and exchange of mankinds main source of information. Therefore, image processing applications will inevitably involve the human aspects of life and work. Image by the access and storage, handling and the impact of various disturbances, indicating when the noise will appear on the screen. In order to reduce the negative impact of noise, clean as much as possible to restore the true picture, we need to use noise reduction filter for image data for processing. Image There are many types of noise such as additive noise, multiplicative noise, salt and pepper noise, Gaussian noise.Common method of noise reduction in the space domain, that is, air filter; also the image data after Fourier transformation to frequency domain, that is, frequency domain filtering method. The main contents of this paper is a frequency domain filtering in image processing applications, frequency domain filtering method is divided into low-pass filtering in frequency domain and frequency domain high-pass filtering method, the following highlights some of the frequency domain low-pass filter and high-pass filter characteristics, and the use of VC + + program, run through the process we can compare the results of several filter images recognition. Key words : Image processing； Noise reduction filter； Low-pass filtering in frequency domain method； High-pass filtering in frequency domain method尽可能地还原干净真实的画面图像识别可能是以图像的主要特征为基础的。每个图像都有它的特征,如字母A有个尖，P有个圈、而Y的中心有个锐角等。对图像识别时眼动的研究表明，视线总是集中在图像的主要特征