Welcome to supertracker¶

An easy-to-use library for implementing various multi-object tracking algorithms.

Overview¶

Supertracker provides a unified interface for multiple object tracking algorithms, making it easy to implement and switch between different tracking approaches in your computer vision applications.

Available Trackers¶

ByteTrack¶

High-performance multi-object tracking
Robust occlusion handling
Configurable parameters for different scenarios
Ideal for real-time applications

Coming Soon¶

DeepSORT: Deep learning enhanced tracking
SORT: Simple online realtime tracking
OCSORT: Observation-centric SORT
BoT-SORT: Bootstrap your own SORT

Quick Installation¶

1	`pip install supertracker`

Basic Usage¶

from supertracker import ByteTrack
from supertracker import Detections

# Initialize tracker
tracker = ByteTrack(
    track_activation_threshold=0.25,
    lost_track_buffer=30,
    frame_rate=30
)

# Process detections
tracked_objects = tracker.update_with_detections(detections)

Support¶

License¶

This project is licensed under the MIT License. See the LICENSE file for details.