EdgeEye

A Linux-based real-time vision monitoring system built with C++ and OpenCV.

Overview

EdgeEye is a Linux vision framework designed for real-time camera processing and future AI vision applications.

The project is developed incrementally, starting from camera acquisition and gradually evolving into a complete edge AI vision system.

---

Features

Current features include:

• USB camera capture
• Real-time camera preview
• FPS monitoring
• CMake build system
• Modular C++ project architecture
• Screenshot capture
• video recording

Planned Features

• Multithreaded camera pipeline
• Qt graphical interface
• Serial communication with STM32
• YOLO object detection
• Edge AI inference

---

Project Structure

EdgeEye
│
├── CMakeLists.txt
├── README.md
├── config.json
├── include/
│   ├── camera.hpp
│   └── fps.hpp
│   └── screenshot.hpp
│   └── video.hpp
│
├── src/
│   ├── main.cpp
│   ├── camera.cpp
│   └── fps.cpp
│   └── screenshot.cpp
│   └── video.cpp
│   
└── build/

---

Environment

• Ubuntu
• GCC
• CMake
• OpenCV

---

Build

mkdir build
cd build
cmake ..
make

---

Run

./EdgeEye

---

Version History

v0.5

• Added video recording module
• Press R to start and stop the current camera video recording
• Added modular video recording management with independent class design
• Automatically creates video directory when saving videos

v0.4

• Added screenshot capture module
• Press S to capture and save the current camera frame
• Added modular screenshot management with independent class design
• Automatically creates screenshot directory when saving images

v0.3

• Added FPS counter module
• Display real-time FPS on the camera preview
• Introduced modular performance monitoring

v0.2

• Integrated OpenCV
• Added USB camera capture
• Implemented real-time camera preview
• Added modular camera class

v0.1

• Initial project setup
• C++ project structure
• CMake build system
• Basic executable framework