Golf Swing Analysis Using Lightweight Neural Network Joint Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing golf swing analysis techniques require expensive instruments or separate sensors/markers to recognize a golfer's posture and motion, making them cumbersome and costly.
Innovation Solution
A method using a lightweight artificial neural network model with depthwise and pointwise convolution to detect joints from photographed images, allowing for comparison with a reference swing without additional sensors, enabling accurate golf swing information estimation on mobile devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If expensive instruments or separate sensors/markers are used to recognize golfer's posture and motion, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent replaces mechanical sensors and markers with a vision-based system using deep learning models. The joint detection model processes images to extract posture information, eliminating the need for physical sensors attached to the golfer's body or expensive motion capture instruments.
Solution Approach 2:
The system creates a digital representation (skeleton model) of the golfer's posture from images rather than directly measuring with sensors. The joint detection model generates coordinate information representing key body points, which is then used for swing analysis without requiring physical contact with the golfer.
2Measurement precision
If separate sensors or instruments are used for golf swing analysis, then measurement precision is improved, but ease of operation deteriorates
Solution Approach 1:
The system uses the golfer's own images to perform self-analysis. The joint detection model automatically processes the images to extract posture information, and the swing analysis is performed without requiring the golfer to attach or operate any external sensors or instruments.
Solution Approach 2:
The image processing system serves multiple functions: it detects joints, reconstructs 3D posture, analyzes swing mechanics, and provides feedback. This single vision-based system replaces multiple specialized instruments that would otherwise be needed for comprehensive swing analysis.
3Measurement precision
If conventional image processing techniques are used to reconstruct 3D images and extract spatial coordinates, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent replaces complex multi-camera 3D reconstruction systems with a single or few camera setup combined with deep learning. The joint detection model directly estimates 3D joint coordinates from 2D images, eliminating the need for multiple synchronized cameras and complex spatial reconstruction algorithms.
Solution Approach 2:
The system changes the approach from direct 3D measurement to 2D image analysis with deep learning parameter estimation. The joint detection model learns to map image features directly to joint coordinates, transforming the problem from geometric reconstruction to pattern recognition.
Data Source
AI summary
A method for estimating information on a golf swing is provided. The method includes the steps of: when a photographed image of a user's golf swing is acquired, detecting at least one joint of the user from the photographed image using an artificial neural network model; comparing the user's golf swing and a golf swing of a comparison target with reference to a position of the at least one detected joint; and estimating information on the user's golf swing on the basis of a result of the comparison.