Comment on: "Extended Hough transform for linear feature detection"

被引:8
作者
Chung, Kuo-Liang [1 ]
Chang, Ting-Chin [1 ]
Huang, Yong-Huai [1 ]
机构
[1] Natl Taiwan Univ Sci & Technol, Dept Comp Sci & Informat Engn, Taipei 10672, Taiwan
关键词
Pixels;
D O I
10.1016/j.patcog.2008.12.026
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The extended Hough transform (EHT) method for line detection with an ability to detect any line segment with desired length, formulated by Cha et al. was discussed. The method continued with the gray image with size hw is transferred into the edge map by using the edge detector. The voting process is performed in two 3D Hough spaces, each Hough space is consisted of a set of 2-D Hough planes, to detect line segments. The developed method is also capable of determining the starting point and ending point of each detected line segment. It has also been demonstrated by experiments, and the study aimed to show that after classifying all edge pixels into two types according to the orientations of edge pixels,the execution time required in the voting process of the EHT method can be reduced efficiently. Experimental results indicate that the average practical execution time improvement ratio is 16% and it is very close to the average theoretical execution time improvement ratio.
引用
收藏
页码:1612 / 1614
页数:3
相关论文
共 2 条
[1]   Extended Hough transform for linear feature detection [J].
Cha, J ;
Cofer, RH ;
Kozaitis, SP .
PATTERN RECOGNITION, 2006, 39 (06) :1034-1043
[2]   USE OF HOUGH TRANSFORMATION TO DETECT LINES AND CURVES IN PICTURES [J].
DUDA, RO ;
HART, PE .
COMMUNICATIONS OF THE ACM, 1972, 15 (01) :11-&