基于码率预分配编码技术的高分辨率航测图像压缩

由于需要对大面阵航空CCD相机带来的庞大航测图像数据进行压缩,在研究多种图像压缩算法的基础上提出了一种基于比特位平面编码的码率预分配图像压缩算法(RPCA)。首先将图像进行多级整数小波变换,以去除图像像素之间相关冗余。根据率失真理论并结合各个子带对图像重建质量的重要性原则,编码前事先确定每个子带在总码率一定的情况下各个子带在实际编码中应当分配的码率大小,再利用自适应MQ算术编码对每个子带比特平面进行熵编码,从而得到细致的嵌入式码流。实验仿真结果表明,该RPCA码率分配精准,图像压缩质量与JPEG2000标准相当,且支持无损到有损的任意倍率图像压缩,但复杂度低于JPEG2000标准,适合于硬件的高速实现。

High-resolution Aerial Image Compression Based on Technology of the Rate Pre-allocation

As it is necessary to compress the large aerial image data from the large-array aerial CCD camera, an image compression algorithm named as Rate Pre-allocated Compression Algorithm (RPCA) based on bit plane coding is proposed by studying variety of image compression algorithms. First, the relative redundancy between pixels of the image should be refused by using multi-level wavelet transform to original image. Second, according to the theory of rate distortion and principle which is the more important to image reconstruction in various sub-bands, the number of bit-rate for different sub-bands should be allocated in the process of actual coding that has been computed. Finally, the coefficients bits are coded by adaptive MQ arithmetic coder. So the careful embedded code stream is produced. Experiment results shows that this algorithm rate allocation is accurate, and the quality of image compression is comparable to the JPEG2000 standard, supporting any lossless to loss image compression ratio, but the complexity is far lower than the JPEG2000 standard and suitable for hardware implementation.