摘要
本文针对Hough变换方法在星空背景下弱小目标运动轨迹检测中所存在的短线段被淹没、峰值模糊化和线段无端点信息问题,对Hough变换算法中参数空间量化方式进行了改进,引入非均匀量化思想,提出了一种基于分块的非均匀Hough变换的轨迹检测方法。实验结果表明该方法既继承了传统Hough变换鲁棒的优点,同时又较好地克服了传统Hough变换的不足,从而有效地解决了星空背景图像中弱小运动目标轨迹检测问题。
While using Hough transform in dim and small target trajectory detection, there will be some problems, such as short trajectory may be submerged, peaks be blurred and no information about the lines＇ ends. To solve these problems, improving the algorithm of Hough transform is necessary. So this paper introduces nonuniform quantization method and proposes an algorithm based On nonuniform quantitative Hough transform in blocks segmented. The experiment results show that the nonuniform quantitative Hough transform algorithm proposed inherits the merits of traditional Hough transform, and overcomes the disadvantages mentioned above, and detects dim and small targets efficiently in the image sequence of star field.
出处
《中国体视学与图像分析》
2009年第1期60-66,共7页
Chinese Journal of Stereology and Image Analysis
基金
国家自然科学基金(60472072)
航空科学基金资助项目(20060853010)
关键词
目标检测
轨迹检测
非均匀量化
HOUGH变换
target detection
trajectory detection
nonuniform quantization
Hough transform
作者简介
姜磊（1984-），男，硕士。
通讯联系人：张艳宁，女，1967年出生，博导，ynzhang@nwpu．edu．cn
