R2-ARC Project

This repository is dedicated to a project that aims to create a droid similar to R2-D2. The droid is controlled via Bluetooth Low Energy from an iOS device and features autonomous following capabilities. These capabilities are powered by the TensorFlow Lite API in conjunction with a Google Coral device, such as the M.2 A+E key Accelerator and provides an Object tracker for use with the detected objects.

Installation

1. First, be sure you have completed the setup instructions for your Raspberry Pi 5 with Google Coral Edge TPU. If it's been a while, repeat to be sure you have the latest software.

   Importantly, you should have the latest TensorFlow Lite runtime installed (as per the Python quickstart).

2. Clone this Git repo onto your computer:

   mkdir main-sbc && cd main-sbc
   
   git clone https://github.com/CSULB-CECS-Senior-Design/Main-SBC.git
   

3. Download the models:

   sh download_models.sh
   

   These models will be downloaded to a new folder models.

4. Install the required Python packages:

   bash install_requirements.sh
   

Test Object Detection

Now that you have the models and requirements installed:

1. Connect your video source (e.g. a webcam) to your Raspberry Pi 5.

2. Run the object detection script:

   cd src
   python3 vision.py

   This script will open a window showing the video feed with detected objects outlined.

Note: Now that you have setup your Raspberry Pi 5 with Google Coral Edge TPU, you can use the official Google Coral repository.