Skip to main content

Abstract

2. Abstract​

Kolam Pattern Digitization and Analysis

This project presents a comprehensive Python-based solution that leverages India's cultural heritage by automating the identification of design principles in Kolam patterns such as symmetry, repetition, and geometric transformations.

The system not only recreates traditional designs programmatically but also provides innovative features for cultural preservation and educational applications.

Key Highlights​

  • Digitization of Traditional Kolam Art
    Detects dots, lines, and geometric patterns in Kolam using advanced Deep Learning models.

  • Deep Learning Models Used

    • ResNet50 – for feature extraction.
    • Faster R-CNN – for object detection (dots, lines, curves).
    • Graph Neural Networks (GNNs) – for relational understanding of design structures.

  • Step-by-Step Kolam Creation
    A unique feature of the system is its ability to demonstrate the gradual, step-by-step process of Kolam creation, offering valuable insights into traditional design methodology.

Applications​

  • Cultural Preservation – Digitally archive and analyze traditional Kolam art.
  • Education – Teach mathematical and geometric concepts through Kolam.
  • Innovation – Enable modern applications such as AR/VR, design generation, and cultural AI.