摘要:矢量量化(VQ)技术是一种先进的数据压缩技术,在图像和语音信号的压缩处理等方面应用广泛。矢量量化是一种先将抽样值进行分组,构成欧氏空间里的矢量,紧接着,再对这些矢量去进行量化的方法。在矢量量化中,数据可以用矢量去表示,可以很容易的对数据去进行压缩和量化。矢量量化压缩比较高,在降低误码率方面,也有很好的效果。本课题主要是利用矢量量化的一种方法——LBG算法对语音信号进行量化压缩。LBG算法是一种经典的迭代算法,程序简单,容易理解,且程序容易实现。本课题针对选取的一段语音信号,利用LBG算法,设计矢量量化器,通过编程实现算法的具体功能,并对语音信号进行矢量量化。本实验实现了矢量量化器的基本量化功能,对语音信号进行了量化压缩,但在矢量量化器的性能改善以及最佳码本的设计方面,仍需进行进一步的研究与改善。
关键词:矢量量化;压缩;LBG算法
Abstract:Vector quantization (VQ) is an advanced data compression technology, which is widely used in image and speech processing.Vector quantizationis a method to quantify the number of vectors in the Euclidean space,And then to quantify the vectors.In vector quantization, the data can be representde by vectors, it is easy to compress and quantify the data, Vector quantization copression is relatively high, and also also has a good effect in reducing the bit error rate. This topic mainly uses one kind of the vector quantization methods---the LBG algorithm , to compress the speech signal. LBG algorithm is a classical iterative algorithm,the program is simple and easy tounderstand, and the program is easy to implement. In this paper, we design a vector quantizer for the selected speech signal, and use the LBG algorithm to realize the function of the algorithm.The algorithm is simple, easy to understand, and the program is easy to implement. The project realized the basic function of vector quantizer, the speech signal of quantization, but in improving the performance and optimun design of vector quantizer codebook, still need to make further research and improvement.
Keywords: Vector quantization;copression;LBG algorithm
目录
第一章 绪论 1
1.1研究的目的和意义 1
1.2矢量量化的发展过程 1
1.3矢量量化在现代的应用 2
1.4本课题研究内容 3
第二章 矢量量化 4
2.1矢量量化的基本原理 4
2.2失真测度的选择 7
2.2.1失真测度的定义 7
2.2.2失真测度的选择要求 8
2.2.3常见的几种失真测度 8
2.3矢量量化器 9
2.4本章小结 10
第三章 最佳矢量量化器设计及LBG算法 11
3.1最佳矢量量化器 11
3.1.1码书设计 11
3.1.2码字搜索 11
3.1.3码字索引分配 12
3.2最佳矢量量化器的设计方案 12
3.2.1最佳划分 12
3.2.2最佳码书