Trajectory Detection for Infrared Dim Point Target Based on Short Trajectory Detection

ZHAO Fei; LU Huan-zhang

doi:10.3969/j.issn.0372-2112.2014.01.028

您当前的位置：

首页 >

文章列表页 >

Trajectory Detection for Infrared Dim Point Target Based on Short Trajectory Detection

更新时间：2025-07-16

- Trajectory Detection for Infrared Dim Point Target Based on Short Trajectory Detection
- Acta Electronica Sinica Vol. 42, Issue 1, Pages: 175-181(2014)
- 作者机构：
  
  国防科技大学ATR国防科技重点实验室,湖南,长沙,410073
- 作者简介：
- 基金信息：
- DOI：10.3969/j.issn.0372-2112.2014.01.028
  CLC： TN911.73
- Published：2014
- 稿件说明：
移动端阅览
ZHAO Fei, LU Huan-zhang. Trajectory Detection for Infrared Dim Point Target Based on Short Trajectory Detection[J]. Acta Electronica Sinica, 2014, 42(1): 175-181.
DOI：

ZHAO Fei, LU Huan-zhang. Trajectory Detection for Infrared Dim Point Target Based on Short Trajectory Detection[J]. Acta Electronica Sinica, 2014, 42(1): 175-181. DOI： 10.3969/j.issn.0372-2112.2014.01.028.

摘要

为解决虚警率较高情况下红外弱小目标轨迹检测效率下降严重的问题，本文提出一种基于短时轨迹检测的红外弱小目标轨迹检测算法.算法首先在相邻帧间进行短时轨迹检测，然后以短时轨迹为基础建立新的轨迹结构并结合轨迹关联算法完成目标轨迹的检测.文章推导了算法的虚警率和检测概率的关系，证明了本文提出的算法相比单帧检测结合轨迹关联的算法具有更好的检测性能.通过对仿真和实际红外图像序列进行实验表明，本文算法能够在准确检测弱小目标的同时有效的减少无效轨迹的数量，从而提高了算法的执行效率.

Abstract

To improve the efficiency under high false alarms probability conditions

an infrared dim point target trajectory detection algorithm based on short trajectories detection is proposed in this paper.The short trajectories are detected in consecutive frames at first

and then the trajectory association which builds a new trajectory depends on the result of short trajectory is applied to detect the target trajectory.The relationship between false alarms probability and detection probability is deduced

and the detection performance is proved to be better than algorithm based on trajectory association combining with segmentation in single frame.The experimental results indicate that the amount of invalid trajectories can be decreased effectively while the target trajectory is being detected accurately

and the efficiency of target detection is improved.

关键词

Keywords

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

The Detection Algorithm for Moving Point Target Trajectory in Image Sequences

Related Author

ZHANG Bing

LU Huan-zhang

Related Institution

ATR Lab of National University of Defense Technology

⁰