3D Skeleton Model Generation via Multi-Viewpoint Image Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current motion capture techniques using inertial navigation require sensor attachment, hindering operations and visual quality, while markerless methods like OpenPose only detect two-dimensional postures, failing to accurately represent skeleton parts hidden by occlusion and resulting in a lack of stereoscopic effect.
Innovation Solution
An information processing apparatus that estimates a three-dimensional skeleton from images captured from multiple viewpoints and generates a stereoscopic skeleton model by modeling joints in a spherical shape and bones in a cylindrical shape, allowing for markerless motion capture and accurate representation of hidden skeleton parts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If markerless motion capture (OpenPose) is used to avoid sensor attachment, then ease of operation is improved, but measurement precision deteriorates due to inability to accurately detect hidden skeleton parts
Solution Approach 1:
The patent transitions from two-dimensional pose estimation (OpenPose) to three-dimensional skeleton estimation by utilizing images captured from multiple viewpoints. This dimensional change enables the system to infer the spatial positions of skeleton parts even when they are occluded in individual 2D images, thereby improving measurement precision while maintaining markerless operation.
2Device complexity
If two-dimensional posture detection is used, then device complexity is reduced, but manufacturing precision deteriorates due to lack of stereoscopic effect
Solution Approach 1:
The patent enhances two-dimensional posture detection by adding a third dimension (depth) through multi-viewpoint image capture and three-dimensional skeleton estimation. This enables accurate representation of skeleton parts hidden by occlusion and generates stereoscopic skeleton models with improved manufacturing precision while keeping the overall device complexity manageable.
3Measurement precision
If sensor attachment is used for motion capture, then measurement precision is improved, but ease of operation deteriorates due to hindrance in sport or game competition
Solution Approach 1:
The patent extracts and removes the motion sensor from the system, replacing it with a markerless motion capture system that uses multiple cameras to capture images from different viewpoints. This eliminates the hindrance caused by sensor attachment during sports or game competitions while maintaining measurement precision through computational methods.
4Device complexity
If two-dimensional skeleton model is generated, then device complexity is reduced, but measurement precision deteriorates due to inability to represent hidden parts accurately
Solution Approach 1:
The patent generates three-dimensional skeleton models instead of two-dimensional models by utilizing images captured from multiple viewpoints. This enables accurate representation of skeleton parts that are hidden by occlusion in any single view, significantly improving measurement precision while the added complexity is managed through efficient image processing algorithms.
Data Source
AI summary
An information processing apparatus, method, and non-transitory computer-readable storage medium that estimate a three-dimensional skeleton based on images captured from a plurality of viewpoints, and generate a skeleton model by modeling the three-dimensional skeleton. The information processing apparatus, method, and non-transitory computer-readable storage medium further model a joint in a spherical shape or a bone in a cylindrical shape in the three-dimensional skeleton, and set a radius of a sphere of the joint or a radius of a cylinder of the bone according to a distance between any one of the plurality of viewpoints or a virtual viewpoint and the joint or the bone.