【信息技术】【2013.05】基于变换和分形编码的图像压缩方法分析

本文为印度Rourkela国立技术研究所（作者：Archana Deshlahra）的硕士论文，共71页。

图像压缩是从图像中去除冗余信息的过程，因此只能存储基本信息，以减少存储大小、传输带宽和传输时间。通过各种变换技术提取基本信息，使其在不损失图像质量和信息的情况下进行重构。本文采用离散余弦变换（DCT）、离散小波变换（DWT）、混合小波变换（DCT+DWT）和分形编码四种变换方法对图像压缩进行了比较分析。针对上述方法编写了相应的MATLAB程序，结果表明，混合DWT-DCT算法在峰值信噪比（PSNR）和较高压缩比下的视觉感知方面均优于独立的基于JPEG的DCT、DWT算法。流行的JPEG标准广泛应用于数码相机和基于网络的图像传输。作为新的JPEG 2000标准的一部分，小波变换声称可以最小化JPEG图像中出现的一些视觉干扰伪影。首先，它使用更大的块（可选，但通常是1024 x 1024像素）进行压缩，而不是原始JPEG方法中使用的8 x 8像素块（通常会产生可见边界）。分形压缩也显现出了有希望的潜力，并声称能够通过插入超出原始分辨率限制的“真实”细节来放大图像。本文对以上每种方法进行了讨论。

Image compression is process to remove theredundant information from the image so that only essential information can bestored to reduce the storage size, transmission bandwidth and transmissiontime. The essential information is extracted by various transforms techniquessuch that it can be reconstructed without losing quality and information of theimage. In this thesis work comparative analysis of image compression is done byfour transform method, which are Discrete Cosine Transform (DCT), DiscreteWavelet Transform( DWT) & Hybrid (DCT+DWT) Transform and fractal coding.MATLAB programs were written for each of the above method and concluded basedon the results obtained that hybrid DWT-DCT algorithm performs much better thanthe standalone JPEG-based DCT, DWT algorithms in terms of peak signal to noiseratio (PSNR), as well as visual perception at higher compression ratio. Thepopular JPEG standard is widely used in digital cameras and web –based imagedelivery. The wavelet transform, which is part of the new JPEG 2000 standard,claims to minimize some of the visually distracting artifacts that can appearin JPEG images. For one thing, it uses much larger blocks- selectable, buttypically1024 x 1024 pixels – for compression, rather than the 8 X 8 pixelblocks used in the original JPEG method, which often produced visibleboundaries. Fractal compression has also shown promise and claims to be able toenlarge images by inserting ―realistic‖ detail beyond the resolution limit ofthe original. Each method is discussed in the thesis.

1 图像压缩
2 图像表示与变换编码
3 性能测量参数
4 性能评估与仿真结果
5 结论与未来研究工作展望

更多精彩文章请关注公众号：