Point Cloud Decoding Indexing via Rectangle Edge Projection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Arctangent approximate computing in existing technologies fails to accurately approximate angles relative to input coordinates and is not suitable for hardware implementation due to division processing, making it inefficient for point cloud decoding.
Innovation Solution
A point cloud decoding device and method that uses the presence of coordinates on an edge of a projected rectangle to assign indices to vertices, determining a projection plane based on maximum and minimum coordinates, and setting a starting point for clockwise or counterclockwise indexing, reducing processing complexity and hardware requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If arctangent approximate computing is used to give an index to each vertex in Trisoup, then the indexing process can be simplified, but the approximation accuracy deteriorates and the magnitude relationship between output angle and input coordinate value does not match ideal arctangent
Solution Approach 1:
The patent changes the parameter representation from continuous angle values (requiring arctangent computation) to discrete index values based on vertex positions on rectangle edges. This parameter transformation eliminates the need for arctangent approximation while maintaining accurate magnitude relationships, as indices directly reflect spatial positions without trigonometric transformation errors.
2Ease of operation
If division processing is used in approximate computing, then angle calculation can be performed, but the technology becomes unsuitable for hardware implementation
Solution Approach 1:
The patent extracts and removes the division operation from the computation process by using a different indexing approach. Instead of calculating angles through division-based arctangent approximation, the system uses direct coordinate comparison and ordering to assign indices, completely eliminating division processing and making the algorithm hardware-friendly.
Solution Approach 2:
The patent substitutes the mechanical division operation with simpler computational operations such as coordinate comparison, sorting, and direct indexing. This replacement maintains the essential functionality of angle-based vertex identification while using operations that are more suitable for hardware implementation, avoiding the complexity of division circuits.
3Device complexity
If a simple indexing method is used for vertices in Trisoup, then hardware complexity is reduced, but the accuracy of angle representation may be compromised
Solution Approach 1:
The patent transitions from representing vertices using angular information (one-dimensional angle space) to using spatial coordinate information (two-dimensional position space). By projecting vertices onto rectangle edges and using their positional coordinates for indexing, the system achieves accurate representation without requiring complex angle calculation hardware, effectively using an additional spatial dimension to replace angular computation.
Data Source
AI summary
A point cloud decoding device 200 according to the present invention includes: an approximate-surface synthesizing unit 2030 configured to use presence of coordinates of each vertex only on an edge of a rectangle projected onto a plane, and give an index to each of the vertces on a basis of a coordinate value of each of the vertices, wherein the approximate-surface synthesizing unit 2030 sets a starting point on the edge of the rectangle prolected onto the plane, and gives the index to each of the vertices clockwise or counterclockwise from the starting point.