Golf Swing Analysis Using Lightweight Neural Network
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, limiting their accessibility and practicality.
Innovation Solution
A method and device using a lightweight artificial neural network model with depthwise and pointwise convolution to detect key points on a golf club and golfer's joints from a photographed image, estimating the club's posture and golf swing information without additional sensors or instruments.
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 extracts the essential detection function from complex dedicated instruments and implements it within a mobile device using standard camera and neural network components. This eliminates the need for separate expensive sensors and markers while maintaining detection capability through image processing algorithms.
Solution Approach 2:
The patent replaces physical sensors and markers with computational image processing methods. Instead of using mechanical/optical sensors attached to the golfer's body and club, the system uses a camera to capture images and a neural network to detect key points and estimate posture, substituting physical measurement systems with computational ones.
2Measurement precision
If separate sensors or instruments are used for golf swing analysis, then detection accuracy is improved, but ease of operation deteriorates
Solution Approach 1:
The patent makes the mobile device perform multiple functions: it serves as both the camera for capturing swing images and the processing unit for analyzing the swing. This universal approach eliminates the need for separate dedicated instruments, making the system easier to operate by using equipment users already possess.
Solution Approach 2:
The system enables users to perform their own swing analysis using their personal mobile devices without requiring external specialized equipment or expert intervention. The mobile device autonomously captures images, processes them through the neural network, and provides swing feedback, making the service self-sufficient and accessible.
3Measurement precision
If standard convolution is used in neural network model, then detection accuracy is improved, but processing speed and efficiency deteriorate
Solution Approach 1:
The patent segments the convolution operation into two distinct stages: depthwise convolution for processing each channel independently, and pointwise convolution for combining channel information. This segmentation allows optimization of each stage separately, improving overall processing efficiency while maintaining detection accuracy.
Solution Approach 2:
The patent implements dynamic adaptive feature fusion that adjusts the fusion strategy based on the specific characteristics of the input image and detection requirements. This dynamic approach optimizes processing efficiency by adapting the computational complexity to the actual needs of each detection task rather than using a fixed convolution approach.
Data Source
AI summary
A method for estimating information on a golf swing is provided. The method includes the steps of: detecting, when a photographed image of a user's golf swing is acquired, at least one of a plurality of key points for a shaft of a club and a plurality of key points for a face of the club, and at least one joint of the user from the photographed image using an artificial neural network model; and estimating a posture of the club on the basis of at least one of positions of the plurality of key points for the shaft and positions of the plurality of key points for the face, and estimating information on the user's golf swing with reference to the posture of the club and a position of the at least one joint of the user.