Skripsi
Implementation of yolov5 algorithm for classification and detection of car types using deep learning / Muhammad Rifqi Rajwa Syafirly
Abstrak
The development of computer vision and deep learning technologies has opened new opportunities in vehicle classification but faces challenges in design complexity and visual variation of cars. This research focuses on implementing the YOLO (You Only Look Once) V5 algorithm to classify car types aiming to develop a model that is accurate robust and effective in differentiating between car categories. The dataset used in this research sourced from Kaggle consists of 1 000 images across four car categories (convertible double cabin sport and van) with 80% of the data used for training 10% for validation and 10% for testing. The methodology includes data preprocessing training the YOLOv5 model and evaluating the performance using metrics such as accuracy precision recall and F1-score. The findings show that the model can classify car types with 96% accuracy 96.6% precision 97% recall and 97% F1-score demonstrating reliable performance for each car category.