Detección del casco en ciclistas mediante algoritmos de visión artificial

Abstract

Abstract: In the present work, the development of a system is carried out to detect if the cyclist has or does not have a helmet by studying artificial vision algorithms. The proposal arises from high rate of cyclists traveling on bike paths and parks. For the elaboration of the work, the agile Scrum development methodology was used to have better control of the progress. Tools like Roboflow are needed to label the images and of Google Colab, an environment that served for training the model, as well as the Yolo-v5 object detection algorithm, the same one that found the optimal training to carry out the evaluation of the results. Finally, the questionnaire was applied based on SUS (System Usability Scale) to evaluate the usability criteria by users. students who were surveyed, obtaining a score of 85/100, which means that is in an acceptable range.