Nishitsuji Takashi, Hosono Yudai, Kakue Takashi, Shimobaba Tomoyoshi, Ito Tomoyoshi, Asaka Takuya
Opt Express. 2019 Apr 15;27(8):11594-11607. doi: 10.1364/OE.27.011594.
Video holography has attracted attention after its invention in 1947; however, the enormous amount of data involved in recording and transmitting three-dimensional (3D) images remains a serious issue in electro-holography. Majority of the studies that have investigated holography transmission target the system that transmits the 3D images by compressing the holograms created on the distributor side using various compression techniques such as the conventional video compression techniques. However, the importance of the information in frequency space and characteristics, such as the correlation between adjacent pixels and frames, is different in natural images and holograms; therefore, these approaches are not always effective. In this study, we propose an effective electro-holography compression scheme based on the vector quantization of point light sources (PLSs). Instead of directly compressing a hologram, our method compresses and transmits PLSs from the distributor side and generates a hologram on the receiver side. To reduce the computational load that is required for creating a computer-generated hologram (CGH) on the receiver side, a fast CGH calculation technique has been developed for the vector-quantized PLS data based on the lookup tables (LUTs). This reduces the data rate by 76% when compared to that observed in case of uncompressed CGH transmission with 2K resolution and results in a calculation speed that is 1.34 times faster than that obtained using the conventional LUT method.
视频全息术自1947年发明以来便备受关注；然而，在电子全息术中，记录和传输三维（3D）图像所涉及的大量数据仍是一个严峻问题。大多数研究全息术传输的工作都针对通过使用各种压缩技术（如传统视频压缩技术）压缩在分发端创建的全息图来传输3D图像的系统。然而，自然图像和全息图在频率空间中信息的重要性以及相邻像素和帧之间的相关性等特征有所不同；因此，这些方法并不总是有效。在本研究中，我们提出了一种基于点光源（PLS）矢量量化的有效电子全息术压缩方案。我们的方法不是直接压缩全息图，而是从分发端压缩并传输PLS，并在接收端生成全息图。为了减少在接收端创建计算机生成全息图（CGH）所需的计算量，基于查找表（LUT）为矢量量化的PLS数据开发了一种快速CGH计算技术。与未压缩的2K分辨率CGH传输相比，这将数据速率降低了76%，并且计算速度比使用传统LUT方法快1.34倍。
