Immersive media compression using mpeg immersive video standard

The MPEG Immersive Video (MIV) standard, plays an essential role in the coding and representation of immersive media content. This standard, designated as Part 12 within the MPEG-I suite, is invented to support virtual reality and extended reality with Six Degrees of Freedom (6DoF). MIV is designed to work on two formats, Multiview plus depth (MVD) which is the scene captured from multi-cameras arranged in different locations, and Multi Planar (MPI/MSI) which is a representation of a 3D scene that involves multiple layers. MIV standard design to exploit traditional 2D video coding like Advanced Video Codec (H.264), High-Efficiency Video Coding (H.265), and Versatile Video Coding (H.266) which had been developed over two decades. To make it applicable, the input views need to cross multi-steps to be encoded using these 2D techniques. Muti views needed to store efficiently with a new format called Atlases. To make it applicable, the input views need to label as basic views and additional views. Basic views which can be packed as it is in the atlases, while for additional views, a process to delete the redundancy between input views needs to be performed using pruning processes to generate the patches, then these patches will be inserted in Atlases after basic views. And finally, the atlases can be encoded using 2D traditional video coding. MIV provides 3 main profiles to meet diverse user requirements: MIV Main Profile, MIV Extended Profile and Geometry Absent Profile. This research provides an in-depth exploration of the entire MIV workflow, offering both theoretical insights and practical perspectives, with a notable emphasis on the MIV Main Profile. Concluding the research with a crucial phase of quality assessment, which is an essential aspect of any encoding standards evaluation to ensure the alignment with anticipated outcomes.