Browse categories
Explore
Fiverr Pro
English
$
USD
I will build real time sports tracking for live player movement and performance
About this Gig
If you need to know where players are moving during a live game, I can build a real-time sports tracking system that converts video into actionable player data.
I develop computer vision pipelines for player detection, multi object tracking, team identification, player IDs, counting, movement trajectories, heatmaps, and performance metrics.
The system can process live or recorded video and provide real time overlays showing each tracked player, movement trails, labels, and selected metrics. Depending on the sport and camera angle, I can also develop ball tracking, court or field mapping, event detection, pose estimation, and sport specific analytics.
I use technologies such as YOLO, OpenCV, Python, tracking algorithms, pose estimation, and custom deep learning models. The objective is not simply to detect people; it is to maintain reliable player identities and generate useful tracking data that can support coaching, scouting, research, or a sports tech product.
I will define the computer vision workflow, develop the required tracking components, and provide the agreed code, outputs, documentation, and integration guidance.
Programming language:
Python
•
R
•
MATLAB
•
SQL
•
Colab
Frameworks:
Scikit-learn
•
DeepPy
•
Google ML Kit
•
SimpleCV
•
Keras
FAQ
Can you keep the same player ID throughout a match?
Yes, persistent multi object tracking can be implemented, but reliability depends on camera angle, resolution, occlusion, lighting, player similarity, and how often players leave or re enter the scene. I will select and configure the tracking approach around your footage.
Can you calculate player speed or distance covered?
Yes, when the camera setup and calibration provide enough information. Depending on the footage, I can produce movement trajectories, approximate speed, distance, heatmaps, and related metrics.
Can you track both players and the ball?
Yes. Player tracking and ball tracking can be combined in the same pipeline when the ball is sufficiently visible and the footage supports reliable detection.
