Darts Score Detection

Hi, everyone.

I created an application that automatically calculates darts scores by applying object detection. As you can see in the video below, it detects the arrow stuck in the dartboard and displays the score.

This application detects the Bull (the center of the dartboard) and the arrow by transfer learning from SSD-Mobilenet. In general, SSD-Mobilenet can detect objects such as dartboards and arrows, but it is difficult to determine the score. For example, since there are 61 different patterns of dart scores, which are combinations of numbers 1-20 and multiples (Single, Double, Triple) + Bull, we have to make sure that a part of the dartboard is detected accordingly. Therefore, in this project, I created the original dataset and added a neural network to estimate the scores, thus making Darts Score Detection a reality.

For more information, please refer to the link below.

1 Like

Hi @enexsy969323, cool project - thanks for sharing!

Interesting use of detection model to classify the location of the darts. Did you also try detecting the bounding box of the dart board, and determining the score from the relative location of the darts?

Thanks for the comment.
In the beginning, I also tried to detect the bounding box of the dartboard.
However, when an arrow was stuck in the dartboard, the arrow detection did not work, so I used the Bull (center of the dartboard) detection to predict the score.