Skeleton Estimation Error Detection for Gymnastics Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing skeleton estimation techniques fail to accurately detect changes in a human body's skeleton over a series of actions, such as in gymnastics, leading to incorrect scoring due to misinterpretation of complex movements.
Innovation Solution
A skeleton estimation device that analyzes three-dimensional data from a 3D laser sensor to estimate skeletal changes over a predetermined period, using feature amounts and correlation analysis to identify and correct abnormal frames, thereby improving the accuracy of skeletal determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional skeleton estimation techniques are used, then processing speed is maintained, but measurement precision of skeletal changes deteriorates
Solution Approach 1:
The patent segments the skeleton estimation process into distinct functional modules: a skeleton estimation unit that generates initial estimates, a feature amount calculation unit that extracts temporal characteristics, and a defective skeleton detection unit that identifies errors. This segmentation allows each module to specialize in specific tasks, improving overall measurement precision without requiring complete system redesign.
Solution Approach 2:
The patent performs preliminary calculation of feature amounts (such as displacement distances and velocity vectors) before final skeleton determination. By pre-processing skeletal data to extract meaningful features like joint displacement patterns and movement velocities, the system establishes a foundation for more accurate change detection while maintaining manageable complexity through structured data preparation.
2Measurement precision
If visual scoring by multiple persons is used, then scoring reliability is maintained through human judgment, but measurement precision deteriorates due to difficulty in judging sophisticated techniques
Solution Approach 1:
The patent implements feedback mechanisms where the defective skeleton detection unit continuously monitors estimation quality and identifies frames with abnormal skeletal changes. This feedback loop allows the system to detect and correct errors in real-time, improving technique recognition accuracy while maintaining scoring reliability through automated quality control that complements human judgment.
Solution Approach 2:
The patent replaces manual visual scoring with an automated skeleton estimation system that uses 3D laser sensor data and computational algorithms to objectively measure skeletal changes. This substitution eliminates human subjectivity and difficulty in judging complex techniques, providing consistent and precise measurements while maintaining reliability through structured detection methods.
3Measurement precision
If skeleton estimation is performed in units of frames, then productivity is improved through efficient processing, but measurement precision deteriorates due to inability to detect abnormal frames
Solution Approach 1:
The patent extracts specific feature amounts (displacement distances, velocity vectors, acceleration patterns) from the skeletal data to identify defective frames. By separating the detection of abnormal frames from the overall processing flow and using targeted feature extraction, the system maintains high processing efficiency while achieving accurate defective frame detection through focused analysis of critical movement parameters.
Data Source
AI summary
A skeleton estimation device including: memory; and processor circuitry coupled to the memory, the processor circuitry being configured to perform processing including: acquiring estimation information in which a change in a skeleton in a predetermined period is estimated in units of frames; calculating a feature amount of the change in the skeleton in the predetermined period; calculating an approximate line based on the calculated feature amount; calculating a correlation state between the calculated feature amount and the calculated approximate line; and detecting, as a defective skeleton, the frame portion in which estimation is abnormal in the estimation information in the predetermined period based on the correlation state.