节点文献

小波分析在图像匹配中的应用研究

Application of Wavelet Analysis in Image Matching

【作者】 孙暐

【导师】 徐立中;

【作者基本信息】 河海大学 , 控制理论与控制工程, 2002, 硕士

【摘要】 本论文综述了目前国内外图像匹配技术的现状和发展趋势,并对部分匹配方法做出了详细的分析和具体实现。 详细的分析了小波滤波技术和二维信号采样,指出了将传统的图像匹配技术和小波滤波技术结合是解决图像匹配技术的有效途径,同时指出对于二维自然景物五株抽取法也是一种有效的滤波方案。 重点研究了小波滤波技术、分形技术在模板匹配中的应用,以及小波滤波的具体设计。不同的滤波方案,不同的匹配方法,以及结合方法的不同就可以形成不同的匹配算法。在分析匹配方法的同时,提出了一种改进的匹配算法;在分析小波平移性时,改进了小波滤波方法,并做出了仿真试验。本论文还将小波技术用于消噪上,收到了良好的效果,并根据试验结果以及自己的体会作了具体的分析。此外,本论文还讨论了小波与分形有密切的关系。 最后,本论文将面向对象和组件技术应用于在图像处理软件的设计中,并在VC++6.0环境下实现了图像匹配技术。

【Abstract】 The status and the development trend of the image matching is Analyzed in this paper. And some of the matching methods are realized. After Having a research on wavelet analysis and the sampling of two dimension signals, it is pointed out that it is a good method that compounds wavelet method with image matching, simultaneously indicated that quincunix decimation is a good method for natural image. A deep research is focused on the application of wavelet method and fractal geometry in image matching and how to design the program of wavelet, and they are embodied. Adopting different sieve methods, different matching methods, and different integrating methods, we can achieve different arithmetics. At that time, I improve two methods: one is used in image matching, the other is adopted in wavelet analysis. They are implemented and simulated. With deeply analyzing the result, I also apply wavelet analysis to reduce image-noise to achieve a good result. Otherwise the relationship of wavelet analysis and fractal method is discussed. Finally, I apply the technology of oop and com to the software design of image processing. And a good result is achieved in vc++6.0.

  • 【网络出版投稿人】 河海大学
  • 【网络出版年期】2002年 02期
  • 【分类号】TP391.4
  • 【被引频次】4
  • 【下载频次】503
节点文献中: 

本文链接的文献网络图示:

本文的引文网络