Worlds hardest jigsaw vs. puzzle machine (all white)

A video on YouTube. In Science & Engineering, a Krater category.

Summary by Krater

The creator builds an automated system using a robot, custom software, and machine learning to solve a 4,000-piece all-white puzzle. After overcoming issues with edge detection, vibrating tumblers, and data accumulation, the robot successfully assembles the puzzle.

From the video

Answers: How can a robot solve a 4000-piece all-white puzzle?

puzzle solving robot
locality sensitive hashing
edge detection algorithm
custom software development
vibratory bowl tumbler

What it concludes

All-white puzzles are extremely difficult for both humans and robots because there are no pictures on the pieces to match.
Locality-sensitive hashing can be used for reverse image search to find similar edges of puzzle pieces without comparing every edge.
Feeding puzzle pieces into an automated system requires careful handling to avoid cumulative errors and misalignments.

Rate it, review it and add it to your lists in Krater.

Titles and thumbnails from YouTube. Krater isn't affiliated with, endorsed by or sponsored by YouTube or Google.