Video-Based Point Cloud Compression Smoothing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Point clouds generated by sensors like LIDAR systems are large and costly to store and transmit, limiting their use in real-time applications due to high storage and network resource requirements.
Innovation Solution
A system that compresses point cloud data by projecting points onto patch planes, generating patch images with spatial and attribute information, and using video encoding standards like HEVC or AV1 to pack and decode these images efficiently, allowing for real-time transmission and processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If point cloud data is stored and transmitted in its original form, then data accuracy is maintained, but storage cost and transmission time increase significantly
Solution Approach 1:
The patent divides the point cloud data into multiple patches, where each patch represents a localized region of the point cloud. This segmentation allows for independent processing and compression of each patch, reducing the overall data transmission time while maintaining local geometric accuracy. The encoder processes each patch separately, generating compressed representations that can be transmitted more efficiently.
Solution Approach 2:
The patent transforms point cloud data from 3D spatial coordinates into 2D patch images with associated depth maps. This parameter transformation changes the data representation from discrete point coordinates to continuous image formats, enabling the use of efficient 2D video compression standards while preserving the essential geometric information through depth encoding.
2Productivity
If point cloud data is compressed using video encoding standards, then transmission efficiency is improved, but data reconstruction accuracy may deteriorate
Solution Approach 1:
The patent converts 3D point cloud data into a 2D representation by projecting points onto patch planes and generating corresponding depth maps. This dimensionality reduction enables the use of efficient 2D video compression standards while maintaining the ability to reconstruct 3D geometry by combining the 2D patch images with their depth information, thus balancing compression efficiency with reconstruction accuracy.
3Measurement precision
If smoothing filters are applied to point cloud data, then visual quality is improved, but geometric fidelity is lost
Solution Approach 1:
The patent applies smoothing filters during the encoding process before compression, rather than during decoding or post-processing. This preliminary smoothing reduces high-frequency noise that would otherwise require more bits to encode, improving compression efficiency while maintaining visual quality. The smoothing is applied in a controlled manner that preserves overall geometric structure while reducing artifacts.
Data Source
AI summary
A system comprises an encoder configured to compress attribute and/or spatial information for a point cloud and/or a decoder configured to decompress compressed attribute and/or spatial information for the point cloud. To compress the attribute and/or spatial information, the encoder is configured to convert a point cloud into an image based representation. Also, the decoder is configured to generate a decompressed point cloud based on an image based representation of a point cloud. In some embodiments, an encoder applies an attribute transfer function to attributes of an image based representation with out-of-loop smoothing such that a decoder may apply any smoothing filter supported by the decoder without being restricted to using a similar smoothing filter as the encoder.