Solving jigsaw puzzles presents a unique challenge for artificial intelligence through the application of computer vision. We explored different puzzle matching techniques by considering piece edges, internal features, and color to design a hybrid system that serves as a puzzle solving user aid. By uploading images of the scattered pieces and the completed puzzle, the matching algorithm will attempt to find the proper locations for each piece and provide visual cues to the user. The application was tested on a sample of varied puzzles and sizes, where the accuracy was dependent on the complexity of the puzzle features and declined linearly as the puzzle was split into more pieces.

• This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator

• Wheeler, Ryan
• Aguiar, Matthew
• Laforest, Jadon
• Ly, Winnie
• Ayvazian, Jack

Publisher

• Worcester Polytechnic Institute

Identifier

• 48811
• E-project-030222-134508

Keyword

• Jigsaw Puzzle
• Computer Vision
• Algorithms
• Artificial Intelligence
• SIFT
• Python

Advisor

• Whitehill, Jacob Richard

Year

• 2022

Date created

• 2022-03-02

Resource type

• Major Qualifying Project

Major

• Computer Science

Rights statement

• In Copyright