Point Cloud Patch Alignment and Size Determination in Bounding Box
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 significant storage requirements and time-consuming data transfer over networks.
Innovation Solution
A system that compresses point cloud data by projecting points onto patch planes, generating patch images with spatial and depth information, and encoding these images using video encoding techniques like HEVC, along with auxiliary information for efficient decoding and reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If point cloud data is stored and transmitted in original format, then data quality and completeness are preserved, but storage requirements and transmission time increase significantly
Solution Approach 1:
The point cloud is divided into multiple patches based on spatial location and surface normal vector similarity. Each patch is processed independently to generate corresponding image data and depth information, enabling selective compression and transmission of only necessary data portions while maintaining overall data quality.
Solution Approach 2:
The patent creates a compressed representation (copy) of the point cloud by projecting points onto patch planes and generating image data. This copy contains essential spatial and depth information but with reduced data volume, allowing quality-preserving compression through controlled projection and sampling.
2Quantity of substance
If point cloud data is compressed to reduce storage and transmission costs, then storage requirements and transmission time are reduced, but data processing complexity and reconstruction accuracy may deteriorate
Solution Approach 1:
The patent introduces patch planes as intermediary structures between the original point cloud and the compressed representation. These planes serve as mediators that simplify the compression process by providing a structured framework for projecting and sampling point data, thereby reducing processing complexity while maintaining reconstruction accuracy.
Solution Approach 2:
The compression process transforms point cloud parameters (spatial coordinates, surface normals) into image parameters (pixel values, depth maps). This parameter transformation enables efficient compression using standard image processing techniques while maintaining the essential information needed for accurate point cloud reconstruction.
3Speed
If point cloud data is processed in real-time for applications like augmented reality, then response time is improved, but computational resources and processing time increase
Solution Approach 1:
By segmenting the point cloud into patches and processing them in parallel, the system achieves real-time performance. Each patch can be processed independently using GPU computing, enabling concurrent operations that reduce total processing time while maintaining real-time response requirements for augmented and virtual reality applications.
Data Source
AI summary
A system comprises an encoder configured to compress attribute information and/or spatial 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. The encoder generates an occupancy map and may adjust a size or placement of a patch in the occupancy map to reduce or eliminate redundant points caused by down-sampling and up-sampling of the occupancy map.