Share
Export Citation
Automatic Swimmer Counter for Outdoor Swimming Pool
Nurtanio I.
Aip Conference Proceedings
Abstract
Swimming is the third most popular sport in Indonesia. Public swimming pools are an option for people who don’t have private pools. Unfortunately, there are several dangers lurking in public swimming pools. One of them is the emergence of disease because the urine content is too high in pool water. A solution that can be done is to replace pool water regularly, but changing pool water regularly on a schedule is still not efficient if the number of swimmers swimming is small. An application that can calculate the number of swimmers automatically is needed so that the scheduling of pool cleaning can be dynamic and the addition of disinfectants can run optimally. By utilizing deep learning technology, a system can be created that can detect and count the number of swimmers in a swimming pool. The data used in this study were 343 images. This data is then divided into 263 training data and 80 test data. The test data consisted of 4 groups of data according to the swimming pool conditions at the time the data was collected. While the algorithm used is the 3rd generation You Only Look Once (YOLO) algorithm for object detection. In this study, a system accuracy of 94% was obtained for cloudypool in the morningnoon, 87.8% for cloudypool in the afternoon, 84.9% for cloudy pool in bright afternoons, and 96.5% for clear pool in the morning, with an average accuracy of 90.8%.