A common question I anticipated was the complexity. It’s tricky to balance enough detail for the final image to be recognizable, while keeping the dot count manageable for a fun puzzle. The contour simplification algorithm handles this by focusing on areas of high contrast and ignoring fine texture noise.
A technical detail: I spent significant time optimizing the dot sequencing step. It’s not simply linear. For complex shapes, the algorithm dynamically determines the shortest Eulerian path that touches all necessary contour points, ensuring the final numbered puzzle remains logical and solvable for the user.
I know the printable market is saturated, but I believe the focus on hyper-personalization is the necessary differentiator. I'm especially interested in feedback on the algorithm's performance on photos versus line art. Let me know what you think!