Geometry-Based Point Cloud Compression for Temporal Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video compression techniques, such as AVC, AV1, and VVC, rely on spatial and temporal redundancies but struggle to efficiently compress high-resolution video data without significantly reducing quality, and existing sensor data compression methods are inefficient for increasing data rates and file sizes.
Innovation Solution
A compression system that converts two-dimensional arrays of element values into three-dimensional point clouds using geometry-based point cloud compression, allowing for efficient compression and decompression by mapping positions to coordinates and associating values with data points, and utilizing metadata for reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If traditional spatial and temporal redundancy techniques (motion estimation, motion compensation) are used for video compression, then compression is achieved to some extent, but compression efficiency is insufficient for high-resolution video without quality loss
Solution Approach 1:
The patent transforms traditional 2D video frame compression into 3D point cloud compression by adding a temporal dimension. Video frames are converted into 3D point clouds where spatial coordinates (x, y, z) represent position and additional attributes represent temporal information, enabling geometry-based compression that exploits spatial-temporal correlations more effectively than traditional 2D methods
Solution Approach 2:
The patent changes the fundamental representation parameters from 2D pixel grids to 3D point cloud coordinates. By representing video data as points in 3D space with attributes, the system enables the use of geometry-based compression algorithms (like MPEG G-PCC or Google Draco) that operate differently from traditional video codecs, achieving better compression ratios
2Measurement precision
If display resolutions and video resolutions continue to increase, then quality is improved, but transmission data rates and file sizes increase without substantial quality improvement
Solution Approach 1:
The patent addresses high-resolution video compression by transitioning to 3D point cloud representation. This dimensional transformation allows the compression system to exploit geometric regularities and spatial-temporal correlations that are not visible in traditional 2D representations, achieving efficient compression even at high resolutions
Solution Approach 2:
The patent merges multiple video frames into a single 3D point cloud structure, combining spatial information from different frames into a unified geometric representation. This merging allows the compression algorithm to identify and exploit temporal redundancies across frames more effectively, reducing overall data size while maintaining quality
3Loss of substance
If geometry-based point cloud compression is applied to consecutive video frames, then compression efficiency is improved, but the approach is less effective when frames do not comprise the same objects
Solution Approach 1:
The patent applies local quality by performing geometry-based compression selectively on regions or frames that exhibit spatial-temporal consistency. By identifying areas where objects persist across frames and applying compression only to those consistent regions, the system maintains high compression ratios while adapting to varying scene content
Solution Approach 2:
The patent introduces dynamic adaptability by adjusting the compression strategy based on scene consistency. The system can dynamically switch between geometry-based compression (when frames show consistent objects) and alternative methods (when objects change significantly), making the compression approach versatile for different video content
Data Source
AI summary
A compression system is configured to obtain a plurality (231) of two-dimensional arrays (234-239) of element values, e.g. a plurality of video frames. Same positions in different arrays comprise a value of the same element at a different moment. The compression system is further configured to convert the plurality of two-dimensional arrays of element values to a three-dimensional point cloud (232), which comprises a plurality of data points, by mapping the positions of the element values in the plurality of two-dimensional arrays to coordinates of the data points and associating each of the element values with a corresponding data point in the point cloud. The compression system is further configured to apply geometry-based point cloud compression to the three-dimensional point cloud.