Dart Hit Location Detection Using Multi-Camera History Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing dart game apparatuses face challenges in accurately identifying hit locations of dart pins due to noise in various installation environments and the difficulty in securing a large enough dataset for deep learning algorithms.
Innovation Solution
A method and device utilizing a dart pin location identification network model that processes input data from multiple cameras and historical information to determine hit locations, generating and processing bounding box images for precise dart pin location identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If rule-based algorithms are used for image analysis, then the system is simple to implement, but accuracy is difficult to secure due to noise in various installation environments
Solution Approach 1:
The patent replaces rule-based algorithms with a deep learning-based image analysis system that uses neural networks to automatically learn and identify dart pin hit locations. This substitution enables the system to handle noise and varying installation environments effectively while maintaining ease of implementation through automated learning processes.
Solution Approach 2:
The patent changes the approach from fixed rule-based parameters to adaptive parameters learned through deep learning. The system adjusts its identification criteria dynamically based on training data, allowing it to maintain high accuracy across different installation environments and lighting conditions without requiring manual parameter tuning.
2Measurement precision
If deep learning algorithms are used for image analysis, then accuracy can be improved, but it is difficult to secure a large enough dataset and implement algorithms to process multiple dart pins
Solution Approach 1:
The patent segments the image analysis process into distinct stages: detecting dart pin positions, identifying hit locations on the target, and processing multiple pins sequentially. This segmentation allows the deep learning system to handle complex multi-pin scenarios by breaking them down into manageable processing steps, reducing overall system complexity.
Solution Approach 2:
The patent implements preliminary action by pre-processing images to enhance dart pin visibility and pre-training the deep learning model with synthetic or limited real data. This preliminary preparation reduces the need for large datasets during deployment and simplifies the implementation of the full deep learning pipeline for processing multiple dart pins.
3Measurement precision
If images from multiple cameras are processed, then comprehensive hit location data is obtained, but processing complexity and time increase
Solution Approach 1:
The patent merges images from multiple cameras into a single composite view or processes them through a unified deep learning model that handles multi-camera input efficiently. This merging approach reduces processing time by eliminating redundant operations and enables comprehensive hit location identification across multiple angles simultaneously.
Data Source
AI summary
Disclosed is a computer program stored in a computer readable storage medium, the computer program providing a method of identifying a hit location of a dart pin when being executed in one or more control units of a dart game device, the method including: generating first history information in response to a first dart pin throw; obtaining a dart target photographing image in response to a second dart pin throw, wherein the dart target photographing image includes an image of a dart target and a plurality of dart pins hitting the dart target; and determining, by a dart pin location identification network model, hit location information of a dart pin corresponding to the second dart pin throw based on the first history information and the dart target photographing image.