Résumé:
The thesis focused on implementing and enhancing object detection techniques using computer vision. Two main strategies were explored: Convolutional Neural Networks (CNN) and the You Only Look Once (YOLO) approach. The study began by examining the fundamentals of neural networks to gain a better understanding of object detection mechanisms. A custom CNN architecture was then developed and implemented to suit the specific datasets. Additionally, the performance of the proposed model was compared to YOLO through the implementation of YOLOv5 and YOLOv8. This allowed for the evaluation of the effectiveness of the custom approach and analysis of the results obtained from the different models.
