Partial 3D Person Representation Using Segmented Point Cloud Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating 3D models of a person's body dimensions using depth cameras are limited, as they require multiple personnel or specialized scanning areas and cannot accurately account for subject movement, leading to inaccuracies in the generated models.
Innovation Solution
A computer-implemented method that uses a stationary depth camera to segment and register depth data into point clouds for the torso, head, and arm regions, employing algorithms like Interior Closest Point and Joint Registration of Multiple Point Clouds to merge these segments into a partial 3D representation, allowing for accurate modeling despite subject movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple depth cameras or traversing camera are used to capture body dimensions, then measurement completeness is improved, but device complexity and operational difficulty increase
Solution Approach 1:
The patent segments the body into multiple regions (torso, head, arms, legs) and processes depth data for each segment separately. This allows a single stationary camera to capture complete body measurements by dividing the scanning task into manageable segments, avoiding the need for multiple cameras or complex traversing mechanisms.
Solution Approach 2:
The patent employs dynamic registration techniques that can handle subject movement during scanning. The system adapts to movement by dynamically adjusting the registration process, allowing accurate 3D model generation even when the subject moves, thus maintaining measurement completeness without requiring the subject to remain perfectly stationary or using complex multi-camera setups.
2Measurement precision
If multiple depth cameras are deployed around the person, then measurement accuracy is improved, but ease of operation deteriorates due to requiring multiple personnel
Solution Approach 1:
The patent makes a single depth camera perform multiple functions by capturing depth data from various angles as the subject moves or the camera remains stationary while the subject rotates. This universal approach replaces the need for multiple cameras operated by multiple personnel, maintaining measurement accuracy while dramatically improving ease of operation.
Solution Approach 2:
The system allows the subject to participate in the scanning process by rotating or moving themselves, eliminating the need for operators to physically move cameras around the subject. The subject essentially scans themselves, making the process easy to operate without requiring multiple personnel.
3Measurement precision
If the depth camera traverses around the person, then complete body coverage is achieved, but operational complexity increases requiring specialized scanning areas
Solution Approach 1:
Instead of moving the camera around a stationary subject, the patent inverts the approach by keeping the camera stationary and having the subject rotate or move. This inversion achieves complete body coverage while eliminating the need for specialized scanning areas and complex camera traversal mechanisms, greatly simplifying operation.
Solution Approach 2:
The patent adds the temporal dimension to the scanning process by capturing depth data over time as the subject moves through different positions. This transforms a static single-point measurement into a dynamic multi-position scan, achieving complete body coverage with a single stationary camera without requiring specialized scanning environments.
4Ease of manufacture
If subject movement is not corrected during scanning, then processing simplicity is maintained, but measurement precision deteriorates due to movement inaccuracies
Solution Approach 1:
The patent implements feedback mechanisms where the system continuously monitors subject movement during scanning and adjusts the registration process accordingly. By providing real-time feedback about position changes and correcting for them in the 3D model generation, the system maintains measurement precision while keeping the processing approach relatively simple through automated correction algorithms.
Data Source
AI summary
Methods for generating a partial three-dimensional representation of a person are disclosed herein. In one aspect, a computer-implemented method comprises obtaining depth data of the person captured from a stationary depth camera scanning around the person; segmenting the depth data into a first segment; mapping the depth data of the first segment to a plurality of point clouds; performing pairwise registration on the point clouds of the first segment; segmenting the depth data into a second segment; mapping the depth data of the second segment to a plurality of point clouds; performing pairwise registration on the point clouds of the second segment; and merging the registered point clouds of the first and second segments to generate the partial 3D representation of the person.