夏菽兰, 孙明泽, 张炜宇, 赵力. 图像角点亚像素坐标提取研究[J]. 微电子学与计算机, 2015, 32(5): 21-24,30. DOI: 10.19304/j.cnki.issn1000-7180.2015.05.005
引用本文: 夏菽兰, 孙明泽, 张炜宇, 赵力. 图像角点亚像素坐标提取研究[J]. 微电子学与计算机, 2015, 32(5): 21-24,30. DOI: 10.19304/j.cnki.issn1000-7180.2015.05.005
XIA Shu-lan, SUN Ming-ze, ZHANG Wei-yu, ZHAO Li. Research on Sub-pixel Coordinates Extraction of Image Corner[J]. Microelectronics & Computer, 2015, 32(5): 21-24,30. DOI: 10.19304/j.cnki.issn1000-7180.2015.05.005
Citation: XIA Shu-lan, SUN Ming-ze, ZHANG Wei-yu, ZHAO Li. Research on Sub-pixel Coordinates Extraction of Image Corner[J]. Microelectronics & Computer, 2015, 32(5): 21-24,30. DOI: 10.19304/j.cnki.issn1000-7180.2015.05.005

图像角点亚像素坐标提取研究

Research on Sub-pixel Coordinates Extraction of Image Corner

  • 摘要: 在角点检测技术的研究中融入一些改进的CSS算法判别曲率偏大像素点,提出一种新的基于支撑像素点曲线拟合的角点亚像素重定位算法,通过实验证明该算法会在CSS角点判别的基础上更加精确的提取角点坐标值,在精度性能上相对于改进CSS算法提高了17%左右.

     

    Abstract: In the research of corner detection technology, some improved algorithm of Curvature Scale Space (CSS) have been used to detect the discriminant curvature larger pixels. A new algorithm for corner sub-pixel positioning is proposed which is based on support pixel curve fitting. The experimental results show that more accurate angular coordinates can be extracted by using the algorithm on the basis of CSS corner. And the precision performance has been improved by about 17% compared with the CSS algorithm.

     

/

返回文章
返回