基于matlab的数字图像预测压缩编码_毕业设计论文.docVIP

摘 要 随着时代的发展，人们可以通过Internet获取大量的信息。这些信息中包含着大量的图像信息，它们占据了很大的数据量，这给信息的存储和传输带来了极大的挑战。图像压缩的目的就是用尽量少的字节来表示图像，并且要求重建图像具有较好的质量。利用图像压缩, 可以减轻图像存储和传输的负担, 使得图像在网络上实现快速传输和实时处理成为现实。 通常，图像中局部区域的像素是高度相关的，因此可以利用先前像素的有关知识来对当前像素的灰度值进行估计，这就是预测。本文介绍了数字图像的预测压缩编码。首先，将图像分成8×8大小的像素块。接着，对每一像素块均采用AR模型，利用Burg算法确定最佳线性预测系数。然后，通过线性差分方程计算得到预测值，最后对实际像素值和预测值之间的差值量化后进行算术编码。从而实现数字图像的预测压缩编码。 本文采用MATLAB提供的图形用户界面工具对三幅典型的标准灰度图像进行了预测压缩编码仿真，并用客观标准和主观标准综合评价重建图像的质量。 仿真结果表明：重建图像与原始图像几乎没有任何差异，能够满足人们的视觉需求。另外，数据压缩比较高且峰值信噪比均在20dB-40dB之间。因此，采用MATLAB实现数字图像的预测压缩编码是一种较好的压缩编码方法，能够在实际中得到广泛的应用。 关键词：图像压缩，线性预测，算术编码，MATLAB，图形用户界面 Abstract People can obtain a great amount of information from the Internet with the development of times and it includes a great amount of image information. The image information occupies huge data and it gives great challenge to information storage and transmission. The aim of image compression is to donate image with bytes as few as possible. The reconstructed image is of relatively good quality is also required. The burden of image storage and transmission can be alleviated by image compression and it turns fast transmission and real-time process of an image in internet into reality. Generally, the pixel in the partial region of an image is highly correlated. So the correlated knowledge of foregoing pixel can be used to estimate the gray value of current pixel. This is called prediction. The paper introduces the predictive compression coding. Firstly, the image is divided into several sub-images of size 8×8. Secondly, AR model is used to each sub-image and then Burg algorithm is used to determine the optimum linear prediction coefficient. Thirdly, linear differential equation is used to calculate the predictive value. Finally, the difference between real pixel value and predictive value is quantized before doing arithmetic coding. Then the predictive compression coding of a digital image is realized. The paper simulate the predictive compression coding as to three typical and standard gray i