立体元图像阵列的分布式预测编码

Distributed prediction coding for elemental image array

为解决组合立体成像系统中立体元图像阵列的存储和传输问题,采用分布式预测编码的方法对立体元图像阵列进行压缩.利用立体元图像间的像素关系,将立体元图像阵列变换成子图像阵列;根据子图像阵列的规律性变化,采用最优位移预测结合PRISM分布式编码结构对子图像阵列进行分级编码.实验结果表明,所提算法的率失真性能明显优于JPEG编码标准;在立体元图像阵列分辨率一定的情况下,组合立体成像系统的显示分辨率越高,本方法的压缩性能越好,对未来高分辨率显示的组合立体成像系统具有更好的适应能力.

To solve the problem of storage and transmission of elemental image array in the integral imaging, the distributed prediction coding method is used to compress the elemental image array. First, the elemental image array is transformed to sub-image array according to the pixel relationship between the elemental images, then the hierarchical coding of the sub-image array is performed via the optimal displacement prediction combined with PRISM distributed coding structure in line with the regular changes of the sub-image arrays. The experiment results indicate that the method is obviously prior to the JPEG in the rate distortion property, and when the resolution of the elemental image array is fixed, the higher the display resolution of the integral imaging, the better the compression performance of the method, which will make it have better adaptability in the high display resolution integral imaging in the future.