Overview

The Overlay Algorithms Project focuses on pattern-matching techniques using triangle point overlay. It was developed as part of an Algorithm Course in Fall 2020. Alex Chea (myself) presented a mock-up implementation of Groth’s Algorithm for whale identification based on their markings.

Professor's Feedback

• Richard Hashino, Algorithm professor, noted:

  “I still remember this presentation nearly four years later, since it was so unique and so interesting. Alex wrote his own code based on a technique called ‘triangle point overlay’ and explained how this could be applied to whale identification.”

Repository Structure

GitHub Repository: Overlay Algorithms

Core Directories & Files:

• Overlay/img/: Contains sample images used for testing overlay techniques.
• Overlay/overlay.py: Implements the triangle point overlay technique for pattern recognition.
• Triangle-Area-Algorithms/triangles.py: Supports geometric computations related to triangle overlays.

Implementation Details

1. Triangle Point Overlay Technique

• Uses triangular structures to detect patterns in images.
• Enhances accuracy in pattern recognition by focusing on distinct geometric features.

2. Groth’s Algorithm for Whale Identification

• Compares whale markings by overlaying triangular points.
• Matches features across images using point alignment techniques.
• Improves pattern-matching efficiency by reducing computational complexity.

Code Example (Overlay Calculation)

def compute_overlay(points):
    overlay_matrix = []
    for p1, p2, p3 in combinations(points, 3):
        triangle_area = compute_triangle_area(p1, p2, p3)
        overlay_matrix.append((p1, p2, p3, triangle_area))
    return overlay_matrix