ProyectoCacao2024

This project is a system for automated phenotyping of fruits and plants for 3D analysis making use of several sensors, multispectral cameras and visible light cameras. The project consists of two subsystems, the phenotyping chamber that sets the environment for the analysis and the robotic arm that orbits the cameras around the plant.

• Phenotyping Chamber: In charge of the environmental sensors and actuators. Receives the controls from the Robotic Arm's Raspberry Pi to interact with the components (e.g. Light sensors, Light actuators, Temperature, etc.)

• Robotic Arm: In charge of registering the plant in 3D by orbiting the cameras around it and controlling the environment. Sends the commands via a Raspberry Pi 4 Model B, rotates using a Dynamixel AX-12 and is equipped with multispectral Survey3 cameras.

The system has its own Dashboard that can be accessed locally by the raspberry from a screen that displays information about the sensors and actuators as well as a preview of the RGB camera as it orbits around the plant.

Pre-requisites

• Python
• Docker

Instalation

Enable I2C Interface and Serial Hardware on the Raspberry Pi

sudo raspi-config

Interface Options -> I2C -> Enable
Interface Options -> Serial Port -> (Serial shell) Disable -> (Serial hardware) Enable

Install package dependencies

sudo apt update && sudo apt upgrade -y
sudo apt install --no-install-recommends git build-essential libgl1 libglib2.0-0
rm -fr /var/lib/apt/lists/*

Clone the project and enter its root directory

git clone https://github.com/eggseal/camara-fenotipado.git
cd camara-fenotipado

Modify the environment variables to suit your device and preferences

cd ./python/chamber/
cp .env.example .env
nano .env

The project has the following variables

• DASHBOARD_PORT The port in which the dashboard frontend will be hosted
• API_PORT The port in which the API will be hosted

Then modify the frontend configuration file to match your API_PORT

nano ./frontend/config/nginx.conf

Search for the following line and modify the number after the :

proxy_pass http://host.docker.internal:5000/;

Create and install the python environment

cd backend
python -m venv env
source env/bin/activate
pip install --upgrade pip setuptools

Install the external python library for the Dynamixel servo

git clone https://github.com/ROBOTIS-GIT/DynamixelSDK.git ./lib/dynamixel-sdk
cd ./lib/dynamixel-sdk/python
python setup.py install
cd ../../..

Install the rest of the requirements

pip install --no-cache-dir -r requirements.txt
cd ../../..

Execution

Run the python API as a daemon

cd python/chamber/backend
source env/bin/activate
python main.py &

Build and run the Docker container for the frontend as a daemon

cd ..
docker compose up --build -d

Components

The following tables lists all the components used in both the phenotyping chamber and the robotic arm systems.

Phenotyping Chamber

Tag	Component	Manufacturer	Description	Pin	GPIO	Type
DHT_PIN	AM2301	ASAIR	Digital temperature and humidity sensor	25	37	IN
	LTR390	Adafruit Industries	Ultraviolet light sensor board	SDA, SCL	3, 5	I2C
	BH1750FVI	ROHM Semiconductor	Digital ambient light sensor IC	SDA, SCL	3, 5	I2C
	TSL2561	Adafruit Industries	Digital luminosity sensor for infrared light	SDA, SCL	3, 5	I2C
WHITE_LIGHT	White Power LED		High power white colored LED	17	11	OUT
IR_LIGHT	IR Power LED		High power infrared LED	27	13	OUT
UV_LIGHT	UV Power LED		High power ultraviolet LED	22	15	OUT

Robotic Arm

Tag	Component	Manufacturer	Description	Pin	GPIO	Type
	Dynamixel AX-12	ROBOTIS	Smart actuator with fully integrated DC Servo module	TX	8	UART
	RGB Camera		RGB Camera for visible light	USB		USB
RGN_CAMERA	Survey3 RGN	MAPIR	Survey camera for Red, Green and Near-Infrared	24, USB	18	OUT, USB
RE_CAMERA	Survey3 RE	MAPIR	Survey camera for Red-Edge	23, USB	16	OUT, USB
START_BTN	Button		Generic button for single pulse input	5	29	IN
STOP_BTN	Button		Generic button for single pulse input	6	31	IN
DIR_SWITCH	Switch		Generic switch for continous input			IN

Architecture

Schematic