A Prototype Tool for Assisting Stopwatch Calibration with a Data Recording System Based on Imaging Technique

A stopwatch serves as a crucial device utilized across multiple domains, including athletics, medical settings, and metrology, primarily for the purpose of quantifying time intervals. Stopwatch calibration is necessary to ensure accurate measurements and compliance with national or international standards. This study aims to develop a prototype for stopwatch calibration that employs a data recording method through an imaging technique based on the YOLO (You Only Look Once) algorithm. This prototype is designed to press the start and stop buttons automatically on both the test and the standard stopwatches and recording the time measurement results accurately. The prototype mainly consists of a RTC DS3231 as a time keeper, two push-pull solenoids to push stopwatch buttons, a microcontroller to control the solenoid, a webcam to capture images of the stopwatch display and a laptop to process the captured images. To transform the images into numerical data, this study used the YOLO version 8 algorithm and some programming applications. The findings showed that the prototype performed well in assisting stopwatch calibration activity which are shown by small errors and uncertainties in calibration results, and an effective detection system with an accuracy of 96.7%, precision of 0.967, recall of 1, and F1 score of 0.982. Therefore, this prototype is expected to enhance the efficiency and accuracy of stopwatch calibration for operators.