论文部分内容阅读
根据干涉超光谱图像的特点,提出了一种基于图像分类与曲线拟合的干涉超光谱图像数据分解算法,结合内嵌比特平面编码技术实现干涉超光谱图像的压缩。与JPEG2000一样,该算法实现了有损、无损压缩的兼容。将干涉超光谱图像数据分为主干涉区域与非主干涉区域两类,针对主干涉区域提出了一种相似匹配算法,而对非主干涉区域采用经验模式分解和二次曲线拟合方法进行数据分析,两种分析算法结合起来能够有效地对谱线数据进行分解,从而有利于取得更好的压缩效果。仿真结果表明,提出的算法可以使无损压缩的输出码率降低0.2~0.4bit/pixel,而近无损、限失真压缩的重建图像质量相应提高。
According to the characteristics of interference hyperspectral images, an image decomposition and curve fitting algorithm based on image classification and curve fitting is proposed. The compression of the hyperspectral image is achieved by combining with embedded bit plane coding. Like JPEG2000, the algorithm implements lossy, lossless compression compatibility. The interference hyperspectral image data is divided into two types: primary interference region and non-primary interference region. A similar matching algorithm is proposed for the primary interference region. Empirical mode decomposition and quadratic curve fitting method are used for the data of the non-primary interference region Analysis, the combination of the two analysis algorithms can effectively decompose the spectral data, which is beneficial for obtaining better compression effect. The simulation results show that the proposed algorithm can reduce the output bit rate of lossless compression by 0.2 ~ 0.4bit / pixel, while the quality of reconstructed image with limited loss compression is improved.