3D Voxel Encoding Bounding Box Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The encoding efficiency is reduced when the bounding box is not set appropriately in voxel-based encoding methods for three-dimensional data structures like point clouds and meshes.
Innovation Solution
An information processing apparatus and method that generates and uses information related to the quantization of three-dimensional positions of encoding targets, including a generator for quantization information and a restoring unit for decoding, to optimize the setting of bounding boxes and voxels, thereby improving encoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If voxel-based encoding is used for three-dimensional data, then data compression is achieved, but encoding efficiency is reduced when bounding box setting is not appropriate
Solution Approach 1:
The patent changes the parameter representation from absolute three-dimensional coordinates to relative coordinates based on bounding box offsets. Each vertex position is encoded as (x, y, z) = (bounding_box_x + offset_x, bounding_box_y + offset_y, bounding_box_z + offset_z), where offsets are quantized to voxel grid positions. This parameter transformation enables efficient compression while maintaining encoding speed by reducing the dynamic range of coordinate values.
Solution Approach 2:
The patent segments the three-dimensional space into discrete voxel units within a bounding box. By dividing the continuous coordinate space into discrete grid cells (voxels), the encoding system can represent vertex positions using integer offsets rather than floating-point coordinates, achieving compression without sacrificing encoding efficiency.
2Ease of manufacture
If bounding box setting is not optimized, then encoding simplicity is maintained, but encoding efficiency deteriorates
Solution Approach 1:
The patent performs preliminary bounding box determination before the actual encoding process. The system identifies the minimum and maximum coordinates of all vertices to define the bounding box boundaries, then uses this pre-calculated information to normalize all vertex positions. This preliminary action simplifies the encoding process by establishing a fixed reference frame while improving efficiency through coordinate normalization.
Data Source
AI summary
The present disclosure relates to information processing apparatus and method that makes it possible to suppress a reduction in encoding efficiency. Information relating to quantization of a three-dimensional position of an encoding target is generated. For example, the information relating to the quantization includes information relating to a coordinate system to be subjected to the quantization, information relating to a bounding box for normalization of position information of the encoding target, or information relating to a voxel for quantization of position information of the encoding target. In addition, three-dimensional information of the encoding target is restored from a signal string on the basis of the information relating to the quantization of the three-dimensional position of the encoding target. The present disclosure is applicable to, for example, an information processing apparatus, an image processing apparatus, an electronic device, an information processing method, a program, or the like.