面向3D呈现的有损和无损混合深度视频编码

A lossy and lossless depth visual coding for 3D presentation

针对多个深度视频流提出实时的压缩方法和评估方案,采用一种有损无损混合的编码方法,可以在图像质量和压缩率之间提供一种有效的平衡控制。行程编码(RLE)用于无损压缩,用来保存深度图像素的高位;像素低10bits保存在YUV图像的Y通道,直接使用×264编码。实验表明,所提方法可以在平均时间小于12ms的前提下同时编解码多个深度图。在实时传输中,通过动态调节质量控制级别,可以获取4∶1到20∶1的压缩率。在压缩率10∶1的情况下,主观解码3D重构效果与原始图几乎完全一致。

Color and depth （RGBD） sensors,such as the Microsoft Kinect,are used in many research are- as. In the immersive group-to-group telepresence system based on a cluster of RGBD sensors, the color and depth image streams are sent to a remote site over a network connection where they are used for the real-time reconstruction of the captured participants. Depth map coding plays a crucial role in the system that allows distributed groups of users to meet in a 3D shared virtual world. We develop and evaluate different schemes for the real-time compression of multiple depth image streams. Our analysis suggests that the hybrid lossless-lossy compression approach provides a good tradeoff between quality and com- pression ratio. The lossless compression based on run length encoding is used to preserve the information of the highest bits of the depth image pixels. The lowest 10-bits of a depth pixel value are directly enco- ded in the Y channel of a YUV image and encoded by a X 264 codec. Our experiments show that the proposed method can encode and decode multiple depth image streams in less than 12 ms on average. De- pending on the compression level, which can be adjusted during application, we are able to achieve a com- pression ratio ranging from about 4 ： 1 to 20： 1. Initial results indicate that the quality for 3D recon- structions is almost indistinguishable from the original for a compression ratio up to l0 ： 1.