SIA OpenIR  > 光电信息技术研究室
红外序列图像中舰船检测与跟踪算法研究
Alternative TitleResearch on ship detection and tracking algorithm in infrared sequence images
孙健1,2
Department光电信息技术研究室
Thesis Advisor向伟
ClassificationTN911.73
Keyword目标检测 行列均值分割 目标跟踪 Meanshift Harris
Call NumberTN911.73/S96/2017
Pages87页
Degree Discipline模式识别与智能系统
Degree Name硕士
2017-11-28
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract海面舰船目标的自动检测与跟踪是红外成像制导的核心技术之一,本文以某型红外成像与跟踪系统为背景,面向工程应用,开展了红外舰船目标检测与跟踪的算法研究,本课题的研究对于增加系统对舰船的捕获概率、提高系统制导精度都具有一定的实用价值和意义。本文主要工作如下:目标检测方面,通过对海面背景下的舰船红外图像特点的分析,在现有行列均值分割算法的基础上,提出了边缘生长与行列均值分割相结合的分割方法,利用图像的灰度特征与梯度特征联合对感兴趣区域进行分割,可大幅改善舰船目标内部灰度分布不均匀而产生的分割面积偏小的问题,提高了舰船检测的正确率。目标跟踪方面,在研究了MeanShift、粒子滤波和核相关滤波几种主流跟踪算法的基础上,提出了一种融合多特征自适应尺度MeanShift跟踪和Harris角点块相关匹配的跟踪算法,根据有效的Harris角点块相关匹配点对的数量来区分使用适合的算法,提高了在多种不同环境条件下的跟踪精度和稳定性。在融合算法的实现过程中,提出了一种基于平均灰度搜索的目标边界精确提取方法,能够实现目标边界的精确定位,跟踪窗口尺度随目标形状变化自适应调整,提高了多特征自适应尺度MeanShift与Harris角点块相关匹配算法的跟踪精度。本文算法已成功应用于某型红外成像与跟踪系统中,对海上舰船目标的自动检测与跟踪均取得了非常好的效果。
Other AbstractAutomatic detection and tracking of ship targets at sea is one of the core techniques of infrared imaging guidance. In this paper, based on a certain infrared imaging and tracking system, an algorithm for detection and tracking of infrared ship targets is carried out for engineering applications. The research of this subject has a certain practical value and significance for increasing the acquisition probability of the ship and improving the guidance precision of the system.. The main work is as follows: In the aspect of target detection, through the analysis of ship infrared image characteristics in the sea background, a segmentation method combining edge growth with row and column mean segmentation is proposed based on the existing row and column mean segmentation algorithm. It uses image gray feature and gradient feature to segment the region of interest, which can greatly improve the problem of smaller area of segmentation caused by uneven distribution of ship's internal gray level, and improve the accuracy of ship detection. In the aspect of target tracking, After studying several mainstream tracking algorithms of MeanShift, Particle Filter and Kernel Correlation Filters, a tracking algorithm based on multi-feature adaptive scale MeanShift tracking and Harris corner matching is proposed. It distinguishes the appropriate algorithm based on the effective number of Harris corner matching points, and improves the tracking accuracy and stability under various environmental conditions. In the implementation of the fusion algorithm, an accurate extraction method of target boundary based on average gray level search is proposed. It can realize the precise location of the target boundary, adjust the window scale adaptively with the change of the shape of the target, and improve the tracking accuracy of multi-feature adaptive scale MeanShift and Harris corner matching algorithm. This algorithm has been successfully applied to a certain infrared imaging and tracking system, and it has achieved very good results in automatic detection and tracking of ship targets at sea.
Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/21274
Collection光电信息技术研究室
Affiliation1.中国科学院沈阳自动化研究所
2.中国科学院大学
Recommended Citation
GB/T 7714
孙健. 红外序列图像中舰船检测与跟踪算法研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2017.
Files in This Item:
File Name/Size DocType Version Access License
红外序列图像中舰船检测与跟踪算法研究.p(2162KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[孙健]'s Articles
Baidu academic
Similar articles in Baidu academic
[孙健]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[孙健]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.