RealTimeStream

RealTimeStream is a real-time video streaming application that captures video from a camera, serves it via a Flask server, and allows clients to view the live stream. The project is structured to support both host and slave functionalities, enabling efficient video processing and streaming.

Project Structure

📦RealTimeStream
 ┣ 📂model
 ┃ ┣ 📜lane_seg.pt
 ┃ ┗ 📜yolov8s.pt
 ┣ 📂src
 ┃ ┣ 📂host
 ┃ ┃ ┣ 📜flask_video_stream.py
 ┃ ┃ ┗ 📜slave.py
 ┃ ┗ 📂slave
 ┃ ┃ ┣ 📂yolo
 ┃ ┃ ┃ ┗ 📜yolo.py
 ┃ ┃ ┗ 📜video_stream_client.py
 ┣ 📂test
 ┃ ┣ 📜install_gstreamer.sh
 ┃ ┣ 📜test_csi_camera.py
 ┃ ┣ 📜test_usb_camera.py
 ┃ ┗ 📜test_usb_media_dev.py
 ┗ 📜README.md

Features

• Real-Time Video Streaming: Capture video from a camera and stream it over the network.
• Flask Server: Serve the video stream using a Flask-based server.
• Client Viewer: View the live video stream on a client machine.
• Model Integration: Incorporate pre-trained models for tasks like lane segmentation and object detection using YOLOv8.

Prerequisites

• Python 3.8 or higher
• Pip package manager
• Git (optional, for cloning the repository)
• A camera connected to the host machine
• Network connectivity between host and client machines

Installation

1. Clone the Repository

   git clone https://github.com/Ashington258/RealTimeStream
   cd RealTimeStream

2. Set Up a Virtual Environment (Optional but Recommended)

   conda activate -n realtimestream python=3.10  # On Windows: venv\Scripts\activate

3. Install Dependencies

   pip install -r requirements.txt

4. Install GStreamer (For Testing)

   The test/install_gstreamer.sh script is provided to install GStreamer dependencies required for testing.

   cd test
   chmod +x install_gstreamer.sh
   ./install_gstreamer.sh
   cd ..

Usage

Running the Flask Video Stream Server (Host)

1. Navigate to the Host Directory

   cd src/host

2. Run the Slave Server

   The slave.py script starts the Flask server that captures video from the camera and serves it.

   python slave.py

   The server will start on http://192.168.2.225:5000/video_feed.

Running the Video Stream Client

1. Navigate to the Slave Directory

   cd src/slave

2. Run the Video Stream Client

   The video_stream_client.py script connects to the Flask server and displays the live video stream.

   python video_stream_client.py

   A window titled "Received Video Stream" will appear, displaying the live video. Press q to quit.

3. go to the web to see the video stream http://192.168.2.225:5000/video_feed

4. run the yolo to see

Testing

The test directory contains scripts to test various camera inputs and media devices.

• Test CSI Camera

  python test_csi_camera.py

• Test USB Camera

  python test_usb_camera.py

• Test USB Media Device

  python test_usb_media_dev.py

Models

The model directory contains pre-trained models used for video processing:

• lane_seg.pt: Pre-trained model for lane segmentation.
• yolov8s.pt: Pre-trained YOLOv8 small model for object detection.

Ensure these models are correctly referenced in your source code for proper functionality.

Contributing

Contributions are welcome! Please fork the repository and create a pull request with your changes.

TODO

• 修改网络模式，网卡改AP模式，不依赖于路由器
• 笔记本端部署工程
• 训练集获取