Intelligent puzzle algorithm based on cosine distance

The algorithm of jigsaw puzzle is assembled according to the similarity of neighboring blocks,so how to calculate the similarity measure between puzzle blocks is crucial. At present, most of the jigsaw algorithms focus on calculating the similarity of the edge of the blocks. There is very little research on the algorithm of combining the similarity of the fragment content with the edge similarity. Therefore, this paper considers not only the edge similarity but also the content similarity between the jigsaw puzzles . An algorithm for calculating the content similarity score of Cosine Distance between two puzzle pieces is proposed, and the accuracy of the smart puzzle is measured by weighting the edge similarity between the puzzle pieces and combine the content similarity with the edge similarity . Study the importance of content similarity to jigsaw puzzles. This article test the pictures from the internet and I have taken for myself, the experimental results show that the algorithm has a higher accuracy.