Portfolio

Overview:

This project focuses on detecting, tracking, and counting vehicles as they enter and exit a designated zone using the YOLOv8l model. The system operates in real time to identify vehicles and track their movement through specified lines to determine whether they are “In” or “Out” of the area. The model was trained on a dataset of 4680 images across 10 different vehicle classes, ensuring accurate and efficient tracking.

Screenshots:

Key Features:

1. YOLOv8 for Vehicle Detection:

The system employs the YOLOv8l model, which provides fast and accurate detection of vehicles in real time, enabling it to track cars as they enter and exit predefined zones.

2. In/Out Zone Detection:

The system focuses on counting vehicles based on their direction of movement. It tracks whether vehicles are entering (“In”) or exiting (“Out”) a designated zone, allowing for clear traffic flow analysis.

3. Extensive Dataset: Trained on a dataset of 4680 images, the system is capable of detecting 10 different vehicle classes, ensuring versatility in various traffic conditions.

4. Accuracy: The model achieved a mean average precision (mAP) of 0.764, with precision at 0.757 and recall at 0.687. This ensures reliable vehicle detection and accurate tracking.

Category:

Tags:

Links

• GitHub
• YouTube