3D Depth Map Resolution Enhancement for Area of Interest
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating high-accuracy 3D shape data increase processing load and apparatus size and cost due to the need for high-density 3D coordinate data, which is not efficiently addressed by current techniques.
Innovation Solution
An information processing apparatus and method that obtains 3D shape data, generates a depth map, and increases the resolution of specific areas of interest within the depth map to generate high-density point group data efficiently, reducing the need for high-density data across the entire measurement area.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-density 3D coordinate data is obtained to increase measurement precision, then the accuracy of 3D shape data is improved, but the processing load and apparatus size and cost increase
Solution Approach 1:
The patent applies local quality by differentiating the density requirements across different spatial regions. High-density point group data is generated only for the area of interest (e.g., face region) where detailed accuracy is needed, while lower-density data is used for other areas. This is achieved by generating a depth map at low resolution first, then selectively enhancing the resolution in the area of interest to create high-density point data only where necessary, thus reducing overall processing load while maintaining measurement precision in critical regions.
2Measurement precision
If high-density 3D coordinate data is obtained to increase measurement precision, then the accuracy of 3D shape data is improved, but the size and cost of measurement apparatus increase
Solution Approach 1:
The patent introduces an intermediary approach by using a depth map as a intermediate representation. Instead of directly capturing high-density 3D data requiring complex expensive apparatus, the system first generates a depth map from lower-density data, then uses image processing techniques to enhance the resolution in the area of interest. This intermediary depth map allows the system to achieve high measurement precision in critical areas without requiring high-density data capture across the entire field of view, thereby reducing apparatus complexity and cost.
3Measurement precision
If contour lines are deduced with greater accuracy to increase 3D shape data accuracy, then the detection accuracy of planes is improved, but the processing load for point group data generation increases
Solution Approach 1:
The patent applies segmentation by dividing the 3D space into different regions based on the area of interest. The depth map is processed to identify the area of interest (e.g., face region), and then only this segmented region undergoes high-resolution enhancement to generate dense point group data. The rest of the area maintains lower resolution. This segmentation approach reduces the processing load for point group data generation while maintaining plane detection accuracy in the critical area of interest.
Data Source
AI summary
An information processing apparatus for processing three-dimensional (3D) information is disclosed. The apparatus obtains 3D shape data representing a shape of a 3D object and data of an image of a field of view including the 3D object. The apparatus generates a depth map corresponding to the field of view based on the 3D shape data and increases a resolution of an area of the depth map corresponding to an area of interest set for the image. The apparatus generates 3D shape data based on the depth map with a resolution of an area corresponding to the area of interest increased.