Sistema para rastreamento do tempo qualificatório para corridas usando visão computacional e microcontroladores

Abstract

Implementation of a low-cost automated system for vehicle timing and identification, based on the integration of Internet of Things devices and Computer Vision, aims to reduce common failures associated with manual timing in automotive events. The proposed architecture employs the ESP32-S3-CAM microcontroller for image acquisition and the HC-SR04 ultrasonic sensor as a physical trigger, communicating with a Python-based processing server and a Node.js web interface. The optical character recognition method uses the EasyOCR library, combined with image preprocessing techniques for distortion and brightness correction. Tests conducted in controlled environments and on an urban circuit demonstrated the effectiveness of the sensor stabilization protocol and the consistency of data transmission. Experimental validation indicated a maximum time deviation of 0.455 seconds compared to manual measurement, consolidating the prototype as a technically viable and accessible tool for managing qualifying times.