What is it about?
This article is the result of theoretical and experimental research on dart hit detection and scoring algorithm. It introduces algorithmic optimizations, which run on real-time. The paper should be of interest to readers in the areas of automated steel dart scoring, image processing and multi-threaded programming.
Featured Image
Photo by Anastase Maragos on Unsplash
Why is it important?
This paper is a proof that high accuracies in dart detection and scoring algorithms can also be achieved with low-cost configurable cameras that are positioned parallel to the baseline of the dartboard. In addition, parallel programming techniques has been applied, in order to have a real-time processing. Experimental results reveal that on average a dart is detected in 568ms with an accuracy of 99.63%.
Perspectives
This research was a long journey, which started from 84.06% accuracy and 1426ms average processing time per throw, and ended at 99.63% accuracy and 568ms processing time. We are quite sure there is a room for more improvement, provided that new features and rules are introduced. For future research, we plan to port the current algorithm to a Raspberry Pi framework. Additionally, an exciting feature would be to have an auto-calibration option by the software itself. Finally, a machine learning layer can be introduced, where the algorithm can update parameters dynamically when it makes mistakes.
Ervin Domazet
International Balkan University
Read the Original
This page is a summary of: Real-time optical dart detection and scoring algorithm for steel tip dartboards, ICGA Journal, February 2023, IOS Press,
DOI: 10.3233/icg-230214.
You can read the full text:
Contributors
The following have contributed to this page
