matching of each feature to vectors in the database. – SIFT use best-bin-first (Beis & Lowe, 97) modification to k-d tree algorithm.
確定! 回上一頁