Inspiration
What it does
The user enters two images: An image to be reconstructed is uploaded from the local machine and a machine to dissect is found using natural language querying. The two images are then triangulated using image processing techniques. Then, k-means is run to find color centroids in each image. To ensure an equivalent color histogram, weighted bipartite matching is performed to perform a color swap. Then we implement the canonical Wallace-Bolyai-Gerwin scissor congruence algorithm to deform the dissection image into the target image.
Log in or sign up for Devpost to join the conversation.