Sistema de asistencia visual basado en ESP32-CAM

Abstract

The inclusion of computer vision systems in vehicles and vehicle prototypes is currently supported by various research studies and development projects. However, these systems always present a significant challenge, as they are highly robust and complex, network, with image processing performed in Python using OpenCV. The received image is analyzed through segmentation in the HSV color space to distinguish the floor surface from potential obstacles. To improve system robustness, a Region of Interest (ROI) is implemented, along with morphological filtering and contour analysis.