SmartEye

Multi-view web camera IoT system with realtime AI processing.

Running Full System

1. Make sure you have Docker installed.
2. Set the camera configurations in config.env.
3. If you're using Linux, run ./setup.sh to generate the configuration from config.env.
   If you're using Windows, run ./setup_windows.sh to generate the configuration from config.env.
4. Run docker compose up --build to run the system.
5. Press Ctrl+C to stop and run docker compose down to take the system down.

Interface Endpoints

• Web Interface
  • http://localhost:8094
  • Running just Web UI also possible:
    • cd WebInterface
    • python -m http.server 8094
• Internal Endpoints
  • Video Streams
    • Internal video stream redistributor / multi-client broadcast with XHR HTTP JSON endpoint for listing current streams.
    • List of streams & FPS
    • Individual streams:
      • http://localhost:8095/<stream_name>.mjpeg
  • Data API Server
    • Live AI Detection Data
    • In JSON format as HTTP XHR endpoint.
  • Face Register WebSocket API Server
    • One instance for each camera, availble per camera ID.
    • At ws://localhost:8100-81XX with JSON format.
    • Requires AI Processing restart for update.
  • Kafka WebUI
    • For debugging purposes, a Web UI for Kafka is included.
    • Kafka WebUI
    • Kafka is exposed to the host at localhost:29092.
  • Stream local webcam instead of ESP-CAM for testing purposes.
    • Activate local venv
    • Run python stream_usbcam.py

Project Directory Structure

Below is a tree of the project directory structure with an explanation of the purpose of each file/directory:

SmartEye
├── ESP32CAM-Firmware                   (Arduino code for ESP-CAM Firmware)
│
│
├── dockerfiles                         (Docker image build files)
│   ├── AIProcess-API.Dockerfile        (Environment for WS Face API)
│   ├── AIProcess-Data.Dockerfile       (Environment for AI Data API)
│   ├── AIProcess.Dockerfile            (Environment for AI Process)
│   ├── MJPEGStreamer.Dockerfile        (Environment for stream distributor)
│   └── VideoBuffer.Dockerfile          (Environment for ESP-CAM video buffer)
│
├── docker-template                     (Docker template generation from config.env)
│   ├── docker-compose.yml.jinja        (Base Template)
│   └── generate_template.py            (Generator script)
│
│
├── AIProcess-Data                      (AI Processing Data Store)
│   ├── data_dlib                       (Model Weights for face recognition)
│   ├── data_faces                      (Individual faces registration images)
│   └── features.csv                    (Extracted registration facial features)
│
├── AIProcess-Kafka                     (Code for AI processing of camera feeds)
│   ├── VideoProcess                    (Sub-classes for each algorithm)
│   │   ├── Face_read.py
│   │   ├── Face_store.py
│   │   ├── KafkaFace.py
│   │   ├── KafkaHand.py
│   │   └── KafkaPose.py
│   ├── kafkacon.py                     (Class for Kafka connections)
│   ├── main.py                         (main.py for AI processing)
│   ├── Servedata.py                    (main.py for Kafka AI Data API)
│   └── Server.py                       (main.py for AI Face Register WS API)
│
├── MJPEGStreamer                       (Code for stream redistributor)
│   ├── bcast.go                        (Class for one-to-many broadcast)
│   ├── fps.go                          (Class for FPS counting)
│   ├── kafka.go                        (Class for Kafka connection)
│   ├── main.go                         (Main program)
│   └── mjpeg.go                        (Class for MJPEG encoding)
│
├── VideoBuffer                         (Code for ESP-CAM video buffer)
│   ├── env.go                          (Class for getting configuration)
│   ├── kafka.go                        (Class for Kafka connection)
│   └── main.go                         (Main program)
│
├── WebInterface                        (Client code for web interface)
│   ├── assets                          (CSS/JS/IMG assets)
│   │   ├── css
│   │   ├── img
│   │   └── js
│   ├── docs.html                       (Documentation page)
│   ├── Face_register.html              (Face Register Page)
│   ├── index.html                      (Index page)
│   ├── Login.html                      (User login page)
│   └── Register.html                   (User register page)
│
│
├── config.env                          (Whole-system configuration)
├── README.md                           (This file)
├── setup.sh                            (Docker template generation for Linux)
├── setup_windows.sh                    (Docker template generation for Windows)
└── stream_usbcam.py                    (Testing local webcam streamer)

26 directories, 137 files