Layered Point Cloud Transmission for Lower Latency

The point cloud data transmission is divided into multiple layers including a first layer bitstream and a second layer bitstream. The first layer contains essential geometry information for basic reconstruction, while the second layer contains additional refinement information. This segmentation allows progressive transmission where critical data arrives first, reducing latency for time-sensitive applications while maintaining the option to transmit additional quality-enhancing data later.

The patent applies different encoding precision and compression ratios to different regions or levels of the point cloud data. The first layer uses a encoding method optimized for speed and essential geometry preservation, while the second layer uses a more complex encoding method for regions or points that benefit from higher precision. This allows quality improvements where most beneficial without uniformly increasing processing complexity across all data.

The encoding process is segmented into multiple passes or layers. The first layer encoding focuses on capturing essential geometric features using a relatively simple algorithm, while the second layer encoding refines specific aspects using more complex methods. This segmentation allows the system to achieve high quality by combining results from multiple encoding passes rather than using a single complex encoder, enabling modular optimization and reduced overall complexity.

The first layer bitstream is generated and transmitted before the second layer bitstream. This preliminary action allows the receiver to begin processing and reconstructing point cloud data with acceptable quality before receiving additional refinement data. The preliminary encoding establishes a foundation that reduces the computational burden of subsequent refinement operations, as the basic structure is already in place.

The point cloud data is segmented into multiple representation layers with different levels of detail. The first layer provides a coarse but sufficient representation for many applications, while subsequent layers provide progressively finer detail. This segmentation enables applications to process only the necessary level of detail based on their requirements, improving processing efficiency by avoiding unnecessary computation on ultra-high-detail data when lower detail suffices.

The system transmits and processes point cloud data at multiple levels of completeness. The first layer bitstream provides partial but sufficient information for basic functionality, while the second layer adds excessive detail that may not always be necessary. This approach allows receivers to achieve acceptable processing efficiency by using only the first layer when appropriate, while having the option to process the more complete second layer when higher quality is required.