Joint Position Estimation Using 2D Base Image Intersections
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for estimating joint positions from point cloud data or images of the human body struggle to accurately determine body curvature and posture, requiring extensive calculations for constructing reference surfaces.
Innovation Solution
A program that estimates joint points on a target's body surface, sets auxiliary lines in a two-dimensional base image, and identifies intersections with the body surface, reducing computational complexity by analyzing a two-dimensional representation rather than a full 3D model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a reference surface is constructed based on a 3D model to estimate joint positions, then measurement precision is improved, but device complexity increases due to large amount of calculation required
Solution Approach 1:
The patent extracts only the necessary information (joint positions) from the complex 3D body surface data by using simplified geometric relationships. Instead of constructing a complete reference surface, the method extracts joint positions directly from key point coordinates and predetermined geometric relationships, significantly reducing calculation complexity while maintaining measurement precision for joint position estimation.
Solution Approach 2:
The patent segments the complex body surface analysis into discrete joint position estimation tasks. By dividing the body surface into regions defined by joint positions and using separate geometric calculations for each joint, the method reduces the overall computational burden compared to constructing a unified reference surface for the entire body.
2Ease of operation
If joint position is estimated from point cloud data or images, then ease of operation is improved, but measurement precision deteriorates as body curvature and posture cannot be accurately determined
Solution Approach 1:
The patent introduces an intermediary geometric model that bridges the gap between simple joint position estimation and accurate body surface analysis. By using a simplified body surface model with predetermined geometric relationships, the method can infer both joint positions and body curvature characteristics without requiring complex 3D reference surface construction, thus maintaining ease of operation while improving measurement precision.
Data Source
AI summary
A program causes a computer to estimate, based on a result of sensing a surface of the body of a target, a plurality of joint points corresponding to joints in the body of the target; set a first type line segment for at least one of the plurality of joint points in a two dimensional base image that represents the body surface of the target, the first type line segment passing through the joint point; and identify an intersection at which the first type line segment intersects with the body surface of the target in the base image.