【作者】 陈志刚；

【导师】 尹福昌；

【作者基本信息】 长春理工大学 ， 光学工程， 2009， 博士

【摘要】 Contourlet变换作为一种新型的多尺度,多分辨率分析工具,不但具备小波的多分辨率,局部性和临界采样等特性,还具备小波欠缺的多方向性和各向异性,所以Contourlet变换比小波变换更能稀疏地表达二维图像。遥感图像具有信息量大,图像细节丰富,Contourlet变换比小波变换有更丰富的系数来表达图像并且能用较少的系数表示图像的细节,所以将Contourlet变换应用于遥感图像的融合和压缩算法之中。为了更好的实现Contourlet变换在遥感图像的融合和压缩算法中应用,首先,将Contourlet变换应用到遥感图像的预处理之中,包括融合之前的遥感图像去噪和配准,压缩编码之前的遥感图像的增强;然后,将Contourlet变换应用到遥感图像的融合之中,包括多源传感器的灰度图像的融合,多光谱和全色遥感图像的融合以及非采样的Contourlet变换在遥感图像融合中的应用;最后,将Contourlet变换应用到遥感图像的压缩编码之中,详细地论述了基于Contourlet变换的嵌入式零树小波编码方法(EZC),SPIHT方法(CSPIHT)和无表零树编码方法(CLZC)。在每个Contourlet变换的应用的实验中验证了这些算法的有效性。为了满足星载全色图像遥感器和多光谱图像遥感器系统中图像的实时处理,设计了基于多DSP处理器协同工作的硬件系统同时也使基于Contourlet变换的融合和压缩算法得以硬件实现,其处理流程是先是将遥感器摄取的图像进行去噪,融合前的配准,再经过融合,增强处理,最后经过压缩编码并采用激光作为通信载体进行传输。更多还原

【Abstract】 Contourlet transform is a new kind of multi-scale and multi-resolution analysis tool and not only has characteristics of multi-resolution, locality and critical sampling which wavelet has, but also has the characteristics of multiple decomposition directions and anisotropy which wavelets lacking. So Contourlet transform has better sparse representation for 2D-image than wavelet transform. Remote sensing image has plenty of information and abundant image details and Contourlet transform has more plentiful coefficients for image representation and can use less Contourlet coefficients to describe image details than wavelet transform, so Contourlet transform can be applied to fusion and compression for remote sensing image.In order to realize fusion and compression for remote sensing image using Contourlet transform, firstly, Contourlet transform is applied to preprocessing for remote sensing image including de-noising and registration before fusion and enhancement before compression for remote sensing image, then, Contourlet transform is applied to fusion for image sensing image including fusion of multi-source gray scale remote sensing images, fusion between multi-spectrum and panchromatic remote sensing image and fusion for remote sensing image based on nonsubsampled Contourlet transform, finally, Contourlet transform is applied to compression coding for remote sensing image including EZC, CSPIHT and CLZC. The experimental results show that the algorithm that Contourlet transform is applied to remote sensing image is effective.In order to meet real time image processing for multi-spectrum and panchromatic remote sensing image on satellite, the hardware system based on multi-DSP processor is designed and realize the fusion and compression algorithm based on Contourlet transform. its work flow is de-noising , registration, fusion, enhancement , compression coding for remote sensing image and transmission by laser.更多还原