论文部分内容阅读
提出了一种使用小波子带熵进行比特分配的遥感图像压缩算法。对遥感图像进行小波提升分解后,分析了各高频子带能量百分比及其熵的变化趋势,在此基础上提出了一种新的快速比特分配方法-使用子带熵进行比特分配。然后对各个高频子带进行均匀量化,量化后的数据采用比特平面编码。对最高比特平面只记录该比特平面中非零系数的坐标,其它比特平面采用行程编码和Huffman编码方法进行压缩。实验结果表明,纹理复杂以及相对平坦的遥感图像使用该算法压缩后都可以获得很好的重构图像质量,峰值信噪比均大于34dB,而压缩比则与图像的复杂程度有关。
A remote sensing image compression algorithm using wavelet subband entropy for bit allocation is proposed. After decomposing the remote sensing images by wavelet, the energy percentage of each high frequency subband and the trend of its entropy are analyzed. Based on this, a new fast bit allocation method is proposed, which uses the subband entropy for bit allocation. Then, each high frequency subband is uniformly quantized, and the quantized data is bit-plane coded. Only the coordinates of non-zero coefficients in the bit plane are recorded for the highest bit plane, while the other bit planes are compressed using run length coding and Huffman coding. The experimental results show that both the complex texture and the relatively flat remote sensing image can be reconstructed with this algorithm. The peak signal-to-noise ratio is better than 34dB, while the compression ratio is related to the complexity of the image.