motcpp

Modern C++ Multi-Object Tracking Library

ByteTrack	OC-SORT
BoostTrack	SORT

Features • Installation • Quick Start • Trackers • Benchmarks • Documentation

---

motcpp is a high-performance, production-ready C++ library for multi-object tracking. It provides state-of-the-art tracking algorithms with a clean, modern C++17 API designed for real-time applications.

✨ Features

• 🎯 9 State-of-the-Art Trackers — SORT, ByteTrack, OC-SORT, DeepOC-SORT, StrongSORT, BoT-SORT, BoostTrack, HybridSORT, UCMCTrack
• ⚡ Blazing Fast — Optimized C++ implementation, 10-100x faster than Python equivalents
• 🔧 Easy Integration — Modern CMake, single-header option, vcpkg support
• 🧪 Well Tested — Comprehensive unit tests with >90% code coverage
• 📦 Cross-Platform — Linux, macOS, Windows
• 🔌 Flexible ReID — ONNX Runtime backend for appearance embeddings
• 📊 MOT Benchmark Ready — Built-in evaluation tools for MOT17/MOT20

📊 Benchmarks

MOT17 Ablation Split

Tracker	HOTA↑	MOTA↑	IDF1↑	FPS
SORT	62.4	75.2	69.2	1250
ByteTrack	66.5	76.4	77.6	1100
OC-SORT	64.6	73.9	74.4	850
UCMCTrack	64.0	75.6	73.9	980
BoostTrack	67.5	77.1	79.2	75

_{Evaluation on the second half of the MOT17 training set using YOLOX detections and FastReID embeddings. Pre-generated data available in releases. FPS measured on Intel i9-13900K.}

C++ vs Python Performance

Tracker	C++ (FPS)	Python (FPS)	Speedup
ByteTrack	1100	45	24x
OC-SORT	850	32	27x
StrongSORT	95	8	12x

🚀 Installation

Prerequisites

• C++17 compiler (GCC 9+, Clang 10+, MSVC 2019+)
• CMake 3.16+
• OpenCV 4.x
• Eigen3

Build from Source

git clone https://github.com/Geekgineer/motcpp.git
cd motcpp
cmake -B build -DCMAKE_BUILD_TYPE=Release
cmake --build build -j$(nproc)
sudo cmake --install build

CMake Integration

find_package(motcpp REQUIRED)
target_link_libraries(your_target PRIVATE motcpp::motcpp)

🎮 Quick Start

#include <motcpp/trackers/bytetrack.hpp>
#include <opencv2/opencv.hpp>

int main() {
    // Create tracker
    motcpp::trackers::ByteTrack tracker;
    
    cv::VideoCapture cap("video.mp4");
    cv::Mat frame;
    
    while (cap.read(frame)) {
        // Your detector outputs: [x1, y1, x2, y2, confidence, class_id]
        Eigen::MatrixXf detections = your_detector(frame);
        
        // Update tracker
        Eigen::MatrixXf tracks = tracker.update(detections, frame);
        
        // tracks: [x1, y1, x2, y2, track_id, confidence, class_id, det_index]
        for (int i = 0; i < tracks.rows(); ++i) {
            int track_id = static_cast<int>(tracks(i, 4));
            cv::Rect box(tracks(i, 0), tracks(i, 1), 
                        tracks(i, 2) - tracks(i, 0), 
                        tracks(i, 3) - tracks(i, 1));
            
            cv::rectangle(frame, box, motcpp::BaseTracker::id_to_color(track_id), 2);
        }
        
        cv::imshow("Tracking", frame);
        if (cv::waitKey(1) == 27) break;
    }
    
    return 0;
}

📋 Trackers

Tracker	Type	Speed	Paper
SORT	Motion	⚡⚡⚡⚡⚡	Bewley et al., 2016
ByteTrack	Motion	⚡⚡⚡⚡⚡	Zhang et al., ECCV 2022
OC-SORT	Motion	⚡⚡⚡⚡	Cao et al., CVPR 2023
UCMCTrack	Motion	⚡⚡⚡⚡	Yi et al., AAAI 2024
DeepOC-SORT	ReID	⚡⚡⚡	Maggiolino et al., 2023
StrongSORT	ReID	⚡⚡	Du et al., TMM 2023
BoT-SORT	ReID	⚡⚡	Aharon et al., 2022
BoostTrack	ReID	⚡⚡	Stanojevic et al., MVA 2024
HybridSORT	ReID	⚡⚡	Yang et al., AAAI 2024

Tracker Selection Guide

Need maximum speed?     → SORT, ByteTrack
General purpose?        → ByteTrack, OC-SORT
Heavy occlusions?       → OC-SORT, UCMCTrack
Moving camera?          → UCMCTrack, BoT-SORT
Re-identification?      → StrongSORT, BoostTrack
State-of-the-art?       → BoostTrack

📚 Documentation

Resource	Description
📖 Getting Started	Installation and first steps
🔍 API Reference	Complete API documentation
📝 Tutorials	Step-by-step guides
💡 Examples	Code examples
🎯 Tracker Guide	Choosing the right tracker
🏗️ Architecture	System design
📊 Benchmarking	Run your own benchmarks

🧪 Testing

cmake -B build -DMOTCPP_BUILD_TESTS=ON
cmake --build build
cd build && ctest --output-on-failure

🤝 Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

📜 Citation

If you use motcpp in your research, please cite:

@software{motcpp2026,
  author = {motcpp contributors},
  title = {motcpp: Modern C++ Multi-Object Tracking Library},
  year = {2026},
  url = {https://github.com/Geekgineer/motcpp},
  license = {AGPL-3.0}
}

📄 License

This project is licensed under the GNU Affero General Public License v3.0 - see the LICENSE file for details.

🙏 Acknowledgments

Special Thanks

This project draws inspiration from and builds upon the excellent work of:

Project	Description
BoxMOT	The original Python multi-object tracking library by Mikel Broström. Our C++ implementation follows similar architecture patterns and algorithm implementations.

Tracking Algorithms

Algorithm	Paper	Authors
SORT	Simple Online and Realtime Tracking	Bewley et al., 2016
ByteTrack	ByteTrack: Multi-Object Tracking by Associating Every Detection Box	Zhang et al., ECCV 2022
OC-SORT	Observation-Centric SORT	Cao et al., CVPR 2023
StrongSORT	StrongSORT: Make DeepSORT Great Again	Du et al., TMM 2023
BoT-SORT	BoT-SORT: Robust Associations Multi-Pedestrian Tracking	Aharon et al., 2022
UCMCTrack	UCMCTrack: Multi-Object Tracking with Uniform Camera Motion Compensation	Yi et al., AAAI 2024
BoostTrack	BoostTrack: Boosting the Similarity Measure and Detection Confidence	Stanojevic et al., MVA 2024
HybridSORT	Hybrid-SORT: Weak Cues Matter for Online Multi-Object Tracking	Yang et al., AAAI 2024

Benchmark Data & Tools

Resource	Source	Citation
MOT17 Dataset	MOTChallenge	Milan et al., 2016
YOLOX Detections	YOLOX	Ge et al., 2021
ReID Embeddings	FastReID	He et al., 2020

---

Made with ❤️ by the motcpp community