学术
摘 要:
本文探讨了粒度计算思想和Hough变换之间的联系，结合小波分析中尺度伸缩和平移的思想，提出了一种基于多粒度数据融合的直线检测算法。在Hough变换中，较大的参数离散间隔（△ρ，△θ）会降低检测精度，而较小的离散间隔则导致峰值扩散和伪峰的出现。该算法改善了（△ρ，△θ）的选择问题，基本实现了以各种合适的粒度（△ρ，△θ）对各种粗细宽度直线的检测。另外，和传统的Hough变换相比，该算法具有更高的计算效率。[著者文摘]
文章出处:
《计算机科学》-2007年34卷9期 -213-217页
栏目信息:
分 类 号:
A Multi-granularity Data Fusion-based Algorithm for Line Detection
XIE Wei-Bo, WANG Yong-Chu, ZHENG Yi-Quan （1.Department of Computer Science, Huaqiao University, Quanzhou 362021; 2.College of Mechanical ＆ Automation, Huaqiao University, Quanzhou 362021）
Abstract:
The paper discusses the relationship between granular computing and Hough transform, combining with the scale flexing and shifting in wavelet analysis, a multi-granularity data fusion based algori/hm for line detection has been proposed. In Hough transform, the bigger interval （△ρ,△θ） can bring down the detecting precision, and a smaller leads to peak spread or pseudo peak appearing. Improved on the selection of （△ρ,△θ）, which line detection for various widths in various appropriate granularities （△ρ,△θ） has achieved basically. In addition to, the algorithm is more efficient in computing than traditional Hough transform.[著者文摘]
Key words:
Granular computing, Hough transform, Line detection, Wavelet analysis, Peak spread, Pseudo peak
基金资助:
本文的研究得到福建省自然科学基金（A0540005）的赞助.
更多>>免费文章
cqvip.com