Overview
Discover how to track vehicles and estimate their speed using advanced computer vision techniques in this comprehensive 25-minute tutorial. Explore object detection with YOLO, implement multi-object tracking using ByteTrack, and leverage Roboflow Inference for efficient processing. Learn to filter detections using polygon zones, understand the mathematics behind perspective transformation, and apply it in code. Dive into the calculations for speed estimation and implement them programmatically. Enhance visualizations to create a polished final product. Gain practical skills to improve traffic control systems, monitor road conditions, and analyze vehicle behavior. Access open-source code, documentation, and additional resources to further expand your knowledge in this cutting-edge field of computer vision.
Syllabus
- Intro
- Object Detection
- Multi-Object Tracking
- Filtering Detections with Polygon Zone
- Math Behind Perspective Transformation
- Perspective Transformation in Code
- Math Behind Speed Estimation
- Speed Estimation in Code
- Visualization Improvements
- Final Results
Taught by
Roboflow