Horse Speed Calculation Using Deep Learning and Optical Flow
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting horse speed in equestrian competitions lack accuracy and scientific intelligence, relying heavily on visual observation rather than precise scientific means.
Innovation Solution
A deep learning-based system using a camera to capture video of horses, employing optical flow field calculation, object detection with YOLOv3, and filtering to determine horse speed, adjusting for camera movement and posture variations, and averaging speeds for accurate predictions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional photo finish systems are used, then the system structure is simple, but the measurement precision of horse speed is insufficient
Solution Approach 1:
The patent replaces traditional mechanical photo finish systems with a deep learning-based image processing system. Multiple cameras capture race images, and a server uses neural network models to automatically detect horse positions, calculate speeds, and determine race outcomes, eliminating the need for complex mechanical measurement devices while achieving higher precision.
Solution Approach 2:
The system uses multiple camera copies to capture images from different angles (front, side, rear cameras). These multiple image copies are processed together by the deep learning system to accurately determine horse positions and speeds, replacing the need for single-point mechanical measurement.
2Productivity
If manual image analysis is used, then the device complexity is low, but the productivity of race result determination is reduced
Solution Approach 1:
The deep learning system performs self-service by automatically analyzing race images, detecting horse and rider positions, calculating speeds, and determining race winners without human intervention. The neural network models autonomously process multiple images from different cameras and generate race results, dramatically improving productivity.
Solution Approach 2:
Manual image analysis is replaced with automated deep learning-based image processing. The system uses convolutional neural networks to automatically extract features from race images, calculate horse speeds based on position changes across multiple frames, and determine race outcomes, eliminating time-consuming manual analysis.
3Reliability
If simple photo finish systems are used, then the device complexity is low, but the reliability of race outcome determination is insufficient
Solution Approach 1:
The patent merges data from multiple independent camera systems (front, side, and rear cameras) into a unified analysis framework. The deep learning system integrates images from different angles and time points to cross-validate horse positions and speeds, ensuring accurate and reliable race outcome determination that is more robust than single-camera systems.
Solution Approach 2:
The system uses feedback from multiple camera perspectives to continuously refine position and speed calculations. By comparing data from front, side, and rear cameras across multiple image frames, the deep learning system can correct errors and verify results, significantly improving the reliability of race outcome determination.
Data Source
Figure 1
Figure 2(a)~2(b)
Figure 3(a)~4
AI summary
This invention discloses a horse speed calculation system and method based on deep learning. Through video of horses walking around a competition field before the game as captured by camera, and through optical flow field calculation, object detection based on deep learning, optical flow field filtering, camera speed calculation, use of horse displacement to adjust camera speed, and unit conversion, average speed of a target horse is ultimately obtained. This invention realizes the use of artificial intelligence technology to observe horses in a more scientific way and to predict and calculate the horse speed.