Building Your Own Real-Time Object Detection App: Roboflow(YOLOv8) and Streamlit (Part 9)

How to Streamlit app in Raspberry Pi locally

Introduction:

The Raspberry Pi has emerged as a versatile and affordable platform that empowers developers to bring AI-powered applications closer to the data source. In this Medium post, we explore the deployment of a Streamlit-based object detection application built on the part 4 on a Raspberry Pi 4 with a camera. Leveraging the power of real-time object detection and the convenience of Streamlit’s intuitive user interface in a local server that not only showcases the potential of object detection but also gives us multiple options to use it.

Advantages of Running Locally

One such advantage is the seamless access to the Raspberry Pi’s built-in camera. By harnessing the potential of the local server, we gain the ability to interact with the camera module directly from our computer, opening up a realm of possibilities for remote surveillance, monitoring, and data collection. This proximity-based approach not only reduces latency but also offers greater control over the application’s performance, ensuring a smoother and more responsive user experience.

In the following sections, I will guide you through the step-by-step process of setting up the Raspberry Pi to run our Streamlit object detection app. So, let’s dive into the world of edge computing and uncover the potential that awaits when we bring AI closer to home.

Prerequisites

To follow this tutorial you’ll need a Raspberry Pi 4, a camera for it and all the other standard accessories.

• Raspberry Pi 4 Model B 4GB or +
• Raspberry Pi Camera Module
• Heat sinks and Fan (optional but recommended)
• 5V 3A USB-C Power Supply
• SD card (at least 8gb)

Raspberry Pi 4 Setup

PyTorch only provides pip packages for Arm 64bit (aarch64) so you’ll need to install a 64 bit version of the OS on your Raspberry Pi

You can download the version used in this guide arm64 Raspberry Pi OS from https://downloads.raspberrypi.org/raspios_arm64/images/ and install it via rpi-imager use this tutorial I create to install the specific version.

Note: For the moment I only tried with the version mentioned but I’ll do my test and update this post with the results using the latest versions. Also 32-bit Raspberry Pi OS will not work.

Once that boots and you complete the initial setup you’ll need to edit the /boot/config.txt file to enable the camera.

Use cd .. in the comand line to change folder until you found the boot folder and use sudo nano config.txt to be able to modify this file.

Make the next changes adding and commenting lines. Once done the changes use ctrl + O and ctrl + X so save the changes and close the file.

# This enables the extended features such as the camera.
start_x=1

# This needs to be at least 128M for the camera processing, if it's bigger you can just leave it as is.
gpu_mem=128

# You need to commment/remove the existing camera_auto_detect line since this causes issues with OpenCV/V4L2 capture.
#camera_auto_detect=1

Installing PyTorch and OpenCV

PyTorch and all the other libraries we need have ARM 64-bit/aarch64 variants so you can just install them via pip and have it work like any other Linux system.

pip install torch torchvision 
pip install opencv-python
pip install numpy --upgrade

We can now check that everything installed correctly:

python -c "import torch; print(torch.__version__)"

Now with this we can start installing the missing libraries

pip install ultralytics 
pip install streamlit

After we have everything installed we need to reboot the Raspberry after the boot we can run the command streamlit hello in the terminal to check everything is ready.

Then we can use git clone with my repository to have a local copy from the project.

git clone https://github.com/fullmakeralchemist/raspberrystreamlit

Once we have the local copy we just need to change the folder to the one that have the run.py file and then we can run from the command line using:

streamlit run app.py

In the white interface we can see the Streamlit in Raspberry Pi but the black one is in my laptop we can do this using the local URL when we run the app, copy and paste the url in the browser from your laptop, this will work only if we have the Pi and the laptop connected to the same router.

In this post of deploying a Streamlit object detection app on a Raspberry Pi and harnessing the advantages of running it locally. The Pi proves to be an invaluable tool for democratizing access to AI and machine learning capabilities. Its compact form factor, cost-effectiveness, and versatility make it a powerful device for building and deploying applications that bridge the gap between data sources and intelligent processing. By accessing the Raspberry Pi’s camera through a local server, we unlock the potential for remote monitoring, surveillance, and data acquisition. This proximity-based approach minimizes latency, enhances control, and paves the way for innovative applications across various domains.

Thanks for reading this series of posts. If you find errors following this or feedback about this guide let me know in the comments, thank you for following this post. Good luck with your projects.