论文部分内容阅读
从图像压缩的角度分析了大孔径静态干涉成像光谱仪成像原理 ,并提出了一种新的遥感多光谱图像压缩方案。新方案利用大孔径静态干涉成像光谱仪推扫成像特点提出了一种低存储量 ,帧间小波域匹配的图像序列压缩方法 ,提高图像质量 3～ 4dB。为了保证图像的光谱信息在 8倍压缩比下有效应用 ,系统采用了一种新的感兴趣区域 (ROI)编码技术。感兴趣区域编码时 ,由于方案中采用率失真优化斜率提升 ,而不是比特平面移位 ,从而使图像在相同的光谱分辨率下拥有更好的空间分辨率。实验结果表明 ,算法大大保护了图像的光谱特性 ,在 8倍压缩比下 ,满足了该类干涉多光谱遥感图像的质量要求。
The imaging principle of large aperture static interference imaging spectrometer is analyzed from the aspect of image compression, and a new remote sensing multispectral image compression scheme is proposed. The new scheme utilizes the characteristics of large aperture static interference imaging spectrometer to push out the imaging and proposes a method of image sequence compression with low memory and inter-frame wavelet domain matching to improve the image quality by 3 ~ 4dB. In order to ensure that the spectral information of the image is effectively applied at a compression ratio of 8, the system adopts a new ROI coding technique. When the region of interest is encoded, the slope is optimized due to the adoption of rate-distortions in the scheme, rather than the bit-plane shift, resulting in better spatial resolution of the image at the same spectral resolution. The experimental results show that the algorithm greatly protects the spectral characteristics of the image and satisfies the quality requirements of such interference multispectral remote sensing images at 8x compression ratio.