Region filling is a fundamental operation in computer graphics and image processing. There are broadly two classes of region filling: polygon based and pixel based. The conventional polygon based region filling algorithm typically uses data structures of records and fields. Using these data structures, the region filling process slows down because of the time-consuming operations of records and fields. This paper proposes a new polygon based region filling algorithm by using the proposed data structures of triples. This results in use of more efficient triple operations involving arrays and elements to fill a region. Using data structures of triples, the y-coordinate modification problem that occurs in the conventional algorithm simply disappears. In addition, contrary to the conventional approach, which uses troublesome geometrical considerations in deciding the even number of elements in each linked list, the proposed triple model uses a simple criterion to meet the even number requirement. Most important of all is the fact that the proposed criterion is independent of the polygon geometry. The experimental results strongly support superiority of the proposed algorithm. It is verified that the proposed algorithm is both theoretically and experimentally better than the conventional algorithm.
淡江理工學刊=Tamkang journal of science and engineering 2(4), pp.175-186