Colorimeter Project

A precise and accessible colorimeter designed to identify the predominant hue of unknown samples. Developed as part of an academic project at Universitat Politècnica de Catalunya, this system combines hardware-based measurements and MATLAB-based data processing to deliver reliable color analysis.

Table of Contents

• Introduction
• Features
• Hardware Requirements
• Software Requirements
• How It Works
• Example Results
• License

---

Introduction

The Colorimeter Project is an educational initiative to build a device that measures and identifies the predominant color of an unknown sample. By using an Arduino Leonardo, five LEDs (red, orange, yellow, green, and blue), and a phototransistor. The device calculates color properties, particularly the hue, whilst eliminating effects from brightness and saturation.

This project highlights the integration of hardware and software, utilizing MATLAB to process and analyze data for accurate color identification.

Features

• Accurate Color Detection: Identifies the hue of unknown samples by transforming raw data into hue-only information.
• Compact Design: Utilizes minimal hardware, making it suitable for educational puprposes and for personal projects.
• Calibration System: Ensures consistent results by normalizing measurements against a white reference card.
• MATLAB Integration: Automates data processing for detailed color analysis.

Hardware Requirements

To replicate this project, the following components are required:

• Arduino Leonardo
• 5 LEDs (Red, Orange, Yellow, Green, Blue) with resistors
• 1µF capacitor
• Phototransistor
• Operational amplifier
• Resistors: 470Ω, 8.2kΩ, 10kΩ, 22kΩ, 82kΩ
• Cylindrical light shield (black cardboard for instance)
• Colored cards: white, red, orange, yellow, green, blue, black

See the image below for an example of circuit schematic:

Software Requirements

• Arduino IDE: To upload the control code to the Arduino Leonardo.
• MATLAB: For data normalization, hue calculation, and result interpretation.

The repository includes:

• main.ino: Arduino code for controlling LEDs, reading sensor data, and generating output.
• main.m: MATLAB script for processing the output from Arduino and determining the color.

How It Works

1. Hardware Setup:

   • See the setup in the documentation.
   • The sample is placed under a cylindrical light shield to prevent external light interference.
   • LEDs sequentially illuminate the sample, and a phototransistor records the reflected light.

2. Data Collection:

   • The Arduino code calculates the time taken for a capacitor to charge under each LED's light.
   • Said times are sent to the Serial Monitor.

3. Data Processing:

   • MATLAB normalizes the raw data using the white reference card.
   • It removes brightness and saturation effects, focusing solely on hue.
   • Euclidean distances between the sample and reference colors determine the closest match.

Example Results

Here are some example outputs from the project:

• Input Vectors:

  • White Card: [55, 363, 1354, 1174, 7]
  • Unknown Card: [56, 299, 1650, 2475, 39]

• Output:

  norm_red = 0.6215
  norm_orange = 0.1688
  norm_yellow = 0.5337
  norm_green = 1.9202
  norm_blue = 2.1576
  norm_black = 1.3923
  

The smallest norm corresponds to orange, indicating the sample is closest to this color.

License

This project is licensed under the GNU General Public License v3.0. See the LICENSE file for details.