| 基于矩形扩张的ROI区域标记算法 |
| |
| 引用本文： | 徐波,李坤.基于矩形扩张的ROI区域标记算法[J].南京气象学院学报,2010(6):573-576. |
| |
| 作者姓名： | 徐波 李坤 |
| |
| 作者单位： | 南京林业大学网络中心, 南京 210037;中国科技大学信息科学技术学院, 合肥 230022 |
| |
| 基金项目： | 南京林业大学科研发展基金(X08-300-1) |
| |
| 摘 要： | 为了满足实时性的要求,在运动目标检测与跟踪领域引入感兴趣区域(Region of Interest,ROI)的概念,提出了基于矩形扩张的ROI区域标记算法.通过与传统算法比较发现,矩形张算法具有较高的效率和准确度.实现结果表明:矩形扩张算法能够快速准确地对二值图像中的运动目标区域进行标记,实现运动目标与复杂背景的分离,进而提高运动目标跟踪的效率.
|
| 关 键 词： | 感兴趣区域 二值图像 矩形扩张 主动轮廓 |
| 收稿时间： | 2010/5/4 0:00:00 |
ROI labeling algorithm based on rectangular expansion |
| |
| XU Bo and LI Kun.ROI labeling algorithm based on rectangular expansion[J].Journal of Nanjing Institute of Meteorology,2010(6):573-576. |
| |
| Authors: | XU Bo and LI Kun |
| |
| Institution: | Network Center, Nanjing Forestry University, Nanjing 210037;School of Information Science and Technology, University of Science and Technology of China, Hefei 230022 |
| |
| Abstract: | This paper introduces the idea of ROI (Region of Interest) to moving target detection and tracking research,and proposes the ROI labelling algorithm based on rectangular expansion. The algorithm can label moving targets in binary image quickly and accurately,thus separates moving targets from complex background effectively and improves the efficiency of moving target tracking. Comparison between this ROI labelling algorithm and traditional dynamic clustering algorithm shows that the former can label moving targets with less time complexity and higher efficiency. |
| |
| Keywords: | ROI binaryimage rectangularexpansion activecontours |
|
| 点击此处可从《南京气象学院学报》浏览原始摘要信息 |
| 点击此处可从《南京气象学院学报》下载免费的PDF全文 |
