Abstract
With the development of immersive video, the quality of compressed 3D content has become an important issue. Video-based Point Cloud Compression (V-PCC) is a popular compression method for point cloud sequences; it achieves the highest quality among MPEG proposals. Compressed point clouds suffer from various artifacts when a high quantization parameter (QP) is used. Examining the causes and types of V-PCC artifacts that occur, we propose a framework to remove the highly noticeable outlier and crack artifacts caused by V-PCC so as to improve compressed point cloud visual quality. A subjective experiment showed that our approach provides significantly improved visual quality, and the improvement becomes more obvious with increasing QP values. Objective evaluation with point-to-point Mean Squared Error (p2p-MSE) shows our proposed method can improve point cloud quality and provides competitive results with lower complexity compared with other methods for point cloud outlier removal and inpainting.
Highlights
I MMERSIVE video, an emerging direction of interaction with content, has many applications in augmented and virtual reality
Some of the methods focus on geometry distortion and ignore texture distortion, some methods are not applicable to both static objects and dynamic sequences, and some have poor correlation with human perception
We conduct a subjective test of visual quality; the results indicate our proposed method significantly improves the visual quality of compressed point clouds (PCs) sequences and static objects
Summary
I MMERSIVE video, an emerging direction of interaction with content, has many applications in augmented and virtual reality. The compression of point clouds (PCs), one of the main representation formats for 3D data, has been studied for a long time. MPEG called for proposals [1] to collect ideas for PC compression for three categories: static models, dynamic sequences and dynamic acquisition. Video-based Point Cloud Compression (V-PCC) from Apple Inc. [2] achieves the best subjective and objective quality under a target bit rate for the dynamic sequence category. The artifacts produced by [2] are easy to notice and could strongly influence user experience. Some of the methods focus on geometry distortion and ignore texture distortion ( overall visual quality is affected by both factors), some methods are not applicable to both static objects and dynamic sequences, and some have poor correlation with human perception. We conduct a subjective experiment to validate our approach to artifact removal
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
