COLORS OF PAINTINGS

---

This project implements an asynchronous telegram bot written using the AIOgram library. The business logic consists of extracting the primary colors of the uploaded image and returning the result as an infographic.

Technologies and Tools

Project Description

The problem of determining the primary colors in the image is solved by clustering pixels by the K-Means algorithm. The number of clusters is set by the user (using the pop-up keyboard in the bot). Based on the results of the calculations, a visualization is created, consisting of the transmitted image, the distribution of colors, a list with colors names, and HEX codes. The correctness of these modules is checked with unit tests written using the Pytest library.
The business logic is placed in an asynchronous telegram bot written in the AIOgram framework. The bot is deployed on the Heroku platform.

Project structure

.
├── examples                          # Folder with examples of processed images
│   ├── input_1.png
│   ├── input_2.bmp
│   ├── input_3.jpeg
│   ├── input_4.jpg
│   ├── output_1.png
│   ├── output_2.png
│   ├── output_3.png
│   └── output_4.png
├── images                            # Folder where images will be saved when the bot works
│   ├── input_282515396.jpg
│   └── output_282515396.jpg
├── tests                             # Folder with modules that test the work of the business logic
│   ├── infrastructure                # Sample images for running tests
│   │   ├── image_1.png
│   │   ├── image_2.webp
│   │   ├── image_3.bmp
│   │   ├── image_4.tiff
│   │   ├── image_5.jpeg
│   │   └── image_6.exr
│   ├── test_get_primary_colors.py    # Tests for module get_primary_colors.py
│   └── test_image_creation.py        # Tests for module image_creation.py
├── .gitignore                        # List of files that are not tracked by Git
├── Aptfile                           # File for downloading dependencies on heroku
├── async_business_logic_call.py      # Asynchronous access to the web application's business logic
├── bot.py                            # Logic of the Telegram bot
├── get_primary_colors.py             # Basic calculations for extracting the primary colors of an image
├── image_creation.py                 # Functions for creating infographics
├── nearest_color.pickle              # A serialized file is containing a model that maps a color RGB value to its name
├── nearest_color_by_rgb.py           # Creating nearest_color_by_rgb.py
├── Procfile                          # Specifies the commands that are executed by the app on startup
├── README.md                         # Description of the project
├── requirements.txt                  # All the required external libraries
└── rgb_color_names.csv               # Table in which each line contains RGB color values and the corresponding name

How to Run the Web App

First, you need to create a file called .env in the root directory. The content of the file should be as follows:

BOT_TOKEN = 'your API telegram bot token'
HEROKU_APP_NAME = 'given by Heroku app name' # if you will deploy the bot
PORT = '443'

To launch the bot, run the following commands in the terminal:

pip install -r requirements.txt
python bot.py

How to Use the Web App

If you want to test its work, then type the name of the bot @colors_of_paintings_bot in the search bar of Telegram, click on the search result, click on the Start button and follow the instructions.

You can also open the bot by clicking on the link t.me/colors_of_paintings_bot.

Examples of processed images

Input Images	Output Images