I will analyze sports video with python object detection and player tracking


About this gig
Turn sports footage into useful data with Python computer vision.
I develop video analysis tools for player and ball detection, object tracking, jersey number OCR, and movement metrics. My project combines YOLOX, ONNX Runtime, OpenCV, appearance-based player matching and custom tracking logic, supported by 8,514 manually annotated detections.
Please contact me before ordering with a sample clip, video duration, camera setup, target objects and desired output. We will agree on the scope and deliverables first. Depending on your project, these can include annotated video, tracking results, metrics and Python source code.
Accuracy depends on resolution, lighting, camera movement, occlusion and object size. Jersey numbers must be readable; distance estimates require suitable camera calibration. I do not use facial recognition. GPU performance and real-time processing are not guaranteed.
Custom model training, multi-camera analysis and deployment require a separate scope.
Get to know Natalia
Python Developer, Computer Vision and Video Analysis
- FromUnited States
- Member sinceMar 2026
Languages
Russian, Ukrainian, English
My Portfolio
FAQ
Should I contact you before ordering?
Yes. Send a sample clip, video duration, target objects and desired output. I will assess feasibility and confirm the scope, deliverables and timeline before you order.
Can you guarantee perfect tracking and jersey number recognition?
No. Results depend on resolution, lighting, camera motion and occlusion. Jersey numbers must be readable. I assess a sample first and explain limitations. Distance estimates require suitable camera calibration.
