Shenyang University of Technology/School of Software, Shenyang, China

---

Abstract—A hybrid algorithm which combines multi-point movement, pixel movement and run-length technique is proposed in order to improve the speed of raster converion for circle arc. In the algorithm, the 1/8 circle arc is divided into 4 adjacent regions according to different raster circle pattern, and the method of 4-point movement and runlength I/O, double-point movement and run-length I/O, mixed 2-point and 3-point movement, 2-point movement along diagonal direction are adopted corresponding to different regions respectively. Different strategy is suitable for the feature of pixel pattern in every region. So the calculation numbers of each step is greatly reduced, and I/O operation numbers is decreased also by the use of runlength output. The experiment result shows that the efficiencies are increased 80% and 62% respectively comparing with the classic single-point algorithm and typical double-point algorithm. Compared with the existing run-length algorithm, the derivation of our algorithm is simple, and is made up of basic operations, so it is apt to be implemented by use of hardware.

Index Terms—Circle Drawing, Pattern Analysis, Hybrid Algorithm, Run-Length Algorithm, Raster Conversion

[PDF]

Cite: Haiwen Feng, Lianqiang Niu, Bowen Fu, and Ling Zhong, "A Hybrid Algorithm of Raster Conversion for Circle Based on Pattern Analysis," Journal of Software vol. 7, no. 9, pp. Haiwen Feng, Lianqiang Niu, Bowen Fu, and Ling Zhong, 2012.