Results Already Achieved

Project 1

Abstract

Exploring the Evolution of Color Palettes in Ancient Wall Paintings Across Different Kingdoms Using Machine Learning

Thiyanga S. Talagala, Department of Statistics, Faculty of Applied Sciences, University of Sri Jayewardenepura

Sri Lanka is home to a rich cultural heritage, including vibrant ancient wall paintings. However, limited research has applied data science to analyze these artistic treasures. This study investigates variations in color palettes across different historical kingdoms using machine learning. Digital images of wall paintings were collected from Kandy, Dambulla, and Sigiriya. Hexadecimal color values were extracted from each image to quantify the dominant hues. A supervised machine learning algorithm was then trained to classify paintings based on their respective kingdoms using color patterns. The model demonstrated high accuracy in predicting the kingdoms. This research provides new insights into the artistic evolution of ancient wall paintings.

Graphical Abstract

Methodology Used to Extract Hexadecimal Colour Values

The eyedroppeR in R was used to extract Hexadecimal colour values from the image. An reproducible code is here:

#devtools::install_github("doehm/eyedroppeR")
library(eyedroppeR)
path <- "img/example.jpg"
extract_pal(12, path, label = "Dambulla")

Project 2

Development of an R Package for Creating an Atlas of Mathematically Rooted Artifacts

The development version of the package is available in my github repository at https://github.com/thiyangt/SriLanka

You can install the development version of SriLanka from GitHub with:

# install.packages("devtools")
devtools::install_github("thiyangt/SriLanka")