海天背景下红外小目标检测与轨迹拟合  被引量:2

Infrared small target detection and trajectory fitting under sea-sky background

在线阅读下载全文

作  者:孙权[1] 安博文[1] 

机构地区:[1]上海海事大学信息工程学院,上海201306

出  处:《激光与红外》2016年第11期1419-1423,共5页Laser & Infrared

基  金:国家自然科学基金(No.61171126);上海市重点支撑项目(No.12250501500);交通运输部应用基础研究项目(No.2014329810060)资助

摘  要:在复杂海天背景下机载红外相机对落水人员的搜救中,提出一种针对红外视频中小目标检测高效算法,本算法首先采用形态学滤波对视频预处理,在相对运动背景下采用时间戳的Hough变换去除海天线的影响,更新海面监控区域;其次使用三帧差分分割小目标,采用改进的等价对算法快速计算分割小目标连通区域并标记。最后利用梯度下降法迭代出小目标运动轨迹,预测落水人员位置。实验表明,该算法不仅能有效地去除海天背景达到实时更新运动背景的目的,而且能够高效地检测出运动目标并标记并拟合出曲线方程,在一定时间内可预测小目标出现位置以便及时搜救落水人员。In order to improve the efficiency of maritime search and rescue under complex sea-sky background, a small target detection algorithm in infrared video is proposed. Firstly the video was preproeessed by morphological filtering, and the effects of sea-sky line were removed under dynamic background by Hough transform with timestamp ,then mo- nitoring regions of sea surface were updated. Secondly ,the small targets were segmented by three-frame difference, and connected components were computed and labeled by the improved equivalence pair algorithm. Finally, the trajectory of small target was iterated by gradient descent method to predict the location of the drowning person. The results show that the algorithm can effectively remove the sea-sky background to update moving background real-time and detect moving targets accurately, and it can also label in real-time and fit the curve equation. In a certain time, this method can predict the position of small target so as to rescue the drowning person in time.

关 键 词:形态学滤波 HOUGH变换 等价对标记 轨迹拟合方程 

分 类 号:TP391.4[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象