Raspberry Pi People Detection

Overview

This project utilizes a Raspberry Pi to detect people in real-time using a camera. It leverages OpenCV's HOG (Histogram of Oriented Gradients) object detection and MQTT for message publishing. The system detects people in the camera frame and sends the number of detected people to a specified MQTT broker.

Features

• Real-time people detection using a webcam (Not a pyCam) connected to Raspberry Pi via USB.
• MQTT-based communication to publish the count of detected people.
• Adjustable camera and broker parameters such as index, resolution, and broker configuration via command-line arguments.
• Timestamp logging with ANSI color formatting for better visualization in the terminal.
• .deb package creation for easy installation on Raspberry Pi devices.

Requirements

To run this project, you need:

• Rust: This project is built using the Rust programming language. Ensure you have Rust version 1.82.0 or later installed.

  To install Rust, follow these steps:

  1. Install rustup by running:

  curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh

  2. Once rustup is installed, you can check the Rust version by running:

  rustc --version

  3. If you need to switch to version 1.82.0 (the required version), you can install it using:

  rustup install 1.82.0
  rustup default 1.82.0

• Dependencies: The exact dependencies required for this project will be listed here once identified. However, you may need the following to get started:

  • opencv: OpenCV bindings for Rust.
  • rumqttc: MQTT client for communication.

  You can add dependencies by including them in the Cargo.toml file. Example:

  [dependencies]
  opencv = "0.72"
  rumqttc = "0.13"
  tokio = { version = "1", features = ["full"] }
  chrono = "0.4"

Installation

1. Clone the repository:

   git clone https://github.com/xartokoptiko/raspberrypi_people_detection.git
   cd raspberry-pi-people-detection

2. Install the necessary dependencies:

   cargo build --release

3. Create a .deb package:

   ./package-debian.sh

   This will generate a .deb package in the linux/ directory, which can be installed on a Raspberry Pi with:

   sudo dpkg -i raspberrypi_people_detection_1.0.0-1_amd64.deb

4. Run the program:

   raspberrypi_people_detection [camera_index] [frame_width] [frame_height] [broker_ip] [broker_port]

   Replace the placeholders with the appropriate values. If no arguments are provided, the program will use default values:

   • camera_index: 2
   • frame_width: 1280
   • frame_height: 720
   • broker_ip: 192.168.1.55
   • broker_port: 1883

License

This project is licensed under the Apache License 2.0.