Recently I worked with a Masters student, who later became a good friend, on the idea of using computer vision techniques for bib number identification. The problem domain was marathon runners. He is fond of running in marathons and thought it important to have a method that could be used for automated recognition of runners who have crossed the finish line at a particular moment.
He employed a cascade of algorithms to complete his job. For bib detection, he employed YOLOv4. And then for reading the actual number on the bib, he employed an optical character recognition technique.
His work was published with grace. The best thing I like about it is his expression (he mostly wrote the paper himself). Please read it. You will definitely like it too. And you might as well get interested in the work. Its cool and funky.
This research investigates the use of Convolutional Neural Networks (CNN) and specifically, You Only Look Once ver. 4 (YOLOv4) to detect Racing Bib Numbers (RBNs) in images from running races and then to recognise the actual numbers using Optical Character Recognition (OCR) techniques.
If you found an error, highlight it and press Shift + Enter or click here to inform us.