Title : 
Quadtree traversal algorithms for pointer-based and depth-first representations
         
        
            Author : 
Fuhrmann, Daniel R.
         
        
            Author_Institution : 
Washington Univ., St. Louis, MO, USA
         
        
        
        
        
            fDate : 
11/1/1988 12:00:00 AM
         
        
        
        
            Abstract : 
Simple top-down pointer-based quadtree traversal algorithms for four-neighbor traversal, eight-neighbor traversal, and connected-component labeling are proposed. It is shown that for this class of algorithms there exist efficient pointer-based representations that can require as little as 2N bytes, where N is total number of nodes in the tree. The algorithms can also be implemented using a depth-first quadtree representation, at the cost of additional nonrecursive walks over the data structure to locate the arbitrary sons of interior nodes
         
        
            Keywords : 
computerised pattern recognition; computerised picture processing; trees (mathematics); computerized pattern recognition; computerized picture processing; connected-component labeling; data structure; depth-first representations; nodes; pointer-based representations; quadtree traversal algorithms; Biomedical computing; Costs; Data structures; Displays; Image coding; Image processing; Image reconstruction; Labeling; Testing;
         
        
        
            Journal_Title : 
Pattern Analysis and Machine Intelligence, IEEE Transactions on