【作者】 邓会新；

【导师】 崔宝侠；

【作者基本信息】 沈阳工业大学 ， 系统工程， 2005， 硕士

【摘要】 近十几年来,随着空间技术、电子光学、激光和计算技术的日趋成熟,遥感信息处理技术得到迅速发展。遥感图像的大容量、大信息量和空间高分辨率对于存储和传输速率都提出了极高的要求,因而有效的数据压缩就显得特别迫切和重要。 本文的主要内容是根据遥感图像分辨率高、相关性弱、冗余度小、纹理多的特点,对遥感图像进行分析;结合小波变换、矢量量化、改进遗传算法等方法实现对遥感图像的压缩;并对这些方法在图像压缩中的作用做了较为深入的研究和探讨。 首先对遥感图像压缩的发展概况、压缩技术以及小波变换在图像处理中的应用做了简要介绍。接着,分析了图像小波变换系数的空间和频率分布特点以及矢量量化在图像小波变换中的应用。对遥感图像的统计特性,相关性以及小波变换域统计特性进行研究分析,并得出结论。最后,针对遥感图像的特点,对变换后的图像分别进行不同策略的量化和编码处理,并提出了小波变换与矢量量化相结合的压缩方法,对小波变换结合矢量量化技术的特点进行了分析。重点讨论了矢量量化的关键技术,并在此基础上提出了新的码书设计方法。 在分析和解决遥感图像压缩的一系列问题中,小波多分辨率分析理论以及基于小波变换的矢量量化为其提供了新的思路和方法,较好地解决了数字图像的存储和传输问题,因此有较强的实用性和广泛的应用前景。更多还原

【Abstract】 In recent ten years, with the development of space technique, electron optics, laser and computer science, remote sensing information is developed rapidly. Remote sensing images have a large abundance of information, which needs more space memory and high transmission speed. Therefore, image compression plays a significant role in the field of remote sensing technique.According to the characteristics of remote images, such as high resolution, little correlation, low redundancy, rich texture, etc., remote sensing images were analyzed. Combined wavelet and vector quantization with modified genetic algorithm, .the effective compressive methods are put forward in this paper and the quality of reconstructed images is improved.At first, the development of image compression and the application of wavelet transform are introduced and analyzed. Secondly, the characteristics of wavelet coefficient of image both in space and frequency, and the application of vector quantization in wavelet domain are analyzed. Moreover, contraposing to the statistical characteristics of remote images, relativities of image and statistical characteristics in wavelet domain are analyzed, and the conclusions are drawn. Finally, According to the characteristics of remote images, different kinds of vector quantization and coding are adopted, and a new method of combination of wavelet transform and the technology of vector quantization is applied and analyzed. Especially, the key technique of vector quantization is discussed and three kinds of new design method of codebook are put forward in this paper.To analyze and solve the problems of remote sensing compression, wavelet multi-resolution analysis vector quantity based on wavelet offers new methods. Therefore, the space memory and transmission problem of spatial data are settled better. Besides, requisition for the spatial data of multi-resolution is met. Therefore, they have stronger practicability and extensive application prospect.更多还原