多探测器拼接成像系统实时图像配准 -

姓名
邮箱
手机号码
标题
留言内容
验证码

多探测器拼接成像系统实时图像配准

doi:10.3788/CO.20150802.0211

王新华^{1, 2},
黄玮²,
欧阳继红^1,,

1.
吉林大学计算机科学与技术学院, 吉林长春 130012;
2.
中国科学院长春光学精密机械与物理研究所, 吉林长春 130033

基金项目:国家自然科学基金资助项目(No.61170092);国家自然科学基金重点资助项目(No.61133011)

详细信息

通讯作者:
王新华(1984—)，男，吉林长春人，博士研究生，工程师，2007年、2010年分别于吉林大学获得学士、硕士学位，主要从事计算光学和机器视觉方面的研究。E-mail：xinhuajlu@163.com

中图分类号:TP391.4
计量
- 文章访问数:1805
- HTML全文浏览量:499
- PDF下载量:682
- 被引次数:0
出版历程
- 收稿日期:2014-11-21
- 录用日期:2015-02-23
- 刊出日期:2015-04-25

Real-time image registration of the multi-detectors mosaic imaging system

1.
College of Computer Science and Technology, Jilin University, Changchun 130012, China;
2.
Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, China

摘要

摘要:依据已设计完成的基于同心球透镜的四镜头多探测器阵列拼接成像系统,对该系统图像拼接配准过程所采用的特征检测提取、特征向量匹配与筛选、空间变换模型参数估计等算法进行了研究。首先,采用Fast-Hessian检测子提取参考图像和待配准图像的特征点,并生成加速鲁棒特征(SURF)描述向量。接着,采用快速近似最近邻(FANN)逼近搜索算法获得初始的匹配点对,并对匹配点对特征向量的欧式距离进行排序。然后,参照成像系统光学设计参数设定合理的阈值,筛选并保留下较好的匹配点对。最后,提出了一种改进的渐进式抽样一致性(IPROSAC)算法对空间变换矩阵模型进行参数估计,从而得到参考图像与待配准图像的空间几何变换关系。实验结果表明:该算法对图像尺寸、旋转和光照变化都具有一定的不变性,特征匹配时间为0.542 s,配准变换时间0.031 s,配准误差精度小于0.1 pixel,可以满足成像系统关于图像配准实时性和准确性的要求,具有一定的工程应用价值。
- 探测器阵列/
- 图像拼接/
- 图像配准
Abstract:According to the detector arrays mosaic imaging system designed with four lenses based on concentric spherical lens, its applied algorithms about the image registration is investigated, such as feature detection and extraction, feature vector matching and screening, spatial transformation model and parameter estimation, etc. First, the fast-hessian detection algorithm is used to find features, and generate feature vector of SURF descriptors. Second, the fast approximate nearest neighbor search algorithm is used to obtain the initial matching points and to sort the Euclidean distance between feature vectors in the matching points. Then after screening the feature points, the good ones are preserved based on a reasonable threshold interval from the optical design parameters. Finally, the transform parameters are estimated by using the improved progressive sample consensus method and the spatial geometry transformation relationship is obtained about the reference image and registration image. Experimental results indicate that the algorithm has some invariance about the size, rotation and illumination changes; the feature matching time is 0.542 s, and the registration transform time is 0.031 s; the registration error precision is less than 0.1 pixel, which can meet the requirements of the imaging system about the image registration including good real-time and accuracy performance.
- detector array/
- image mosaic/
- image registration

HTML全文

图 1多探测器拼接成像原理图

Figure 1.Principle diagram of multi-detectors mosaic imaging

下载: 全尺寸图片幻灯片

图 2图像拼接配准算法流程

Figure 2.Procedure of the image registration algorithm

下载: 全尺寸图片幻灯片

图 39×9方框滤波器模板

Figure 3.Box filter with 9×9

下载: 全尺寸图片幻灯片

图 4特征点主方向选取过程示意图

Figure 4.Main direction selection process of the feature points

下载: 全尺寸图片幻灯片

图 5特征描述向量构成示意图

Figure 5.Composition of the feature description vector

下载: 全尺寸图片幻灯片

图 6成像系统的象散、畸变图

Figure 6.Field curve and distortion of the imaging system

下载: 全尺寸图片幻灯片

图 7特征检测测试样本

Figure 7.Test samples of feature detection

下载: 全尺寸图片幻灯片

图 8特征检测提取结果

Figure 8.Result of the feature detection and extraction

下载: 全尺寸图片幻灯片

图 9特征匹配结果

Figure 9.Result of the features matching

下载: 全尺寸图片幻灯片

图 10特征匹配筛选结果

Figure 10.Screen result of the features matching

下载: 全尺寸图片幻灯片

图 11图像配准变换结果

Figure 11.Result of images registration transformation

下载: 全尺寸图片幻灯片

表 1特征检测描述数据对比

Table 1.Data comparison of feature detection and description

图7(a) 图7(b)

下载: 导出CSV

表 2特征匹配数据对比

Table 2.Data comparison of features matching

Brute-Force算法 FANN算法

下载: 导出CSV

表 3配准变换方法与数据对比

Table 3.Methods and data comparison of registration transform

文献[10]方法本文方法

下载: 导出CSV

参考文献 (12)

[1]	[1] MARKS D L,LLULL P R,et al.. Characterization of the AWARE 10 two-gigapixel wide-field-of-view visible imager[J]. Applied Optics,2014,53(13):54-63.
[2]	[2] 史光辉,杨威.用于图像拼接的电视摄像光学系统[J].中国光学,2014,7(4):638-645. SHI G H,YANG W. Optical system used to compose images in television photograph[J]. Chinese Optics,2014,7(4):638-645.(in Chinese)
[3]	[3] 杨磊,曹剑中,唐利孬,等.全景图像自动拼接算法的优化设计[J].红外与工程,2014,43(3):985-990. YANG L,CAO J Z,TANG L A,et al.. Optimized design of automatic panoramic images mosaic[J]. Infrared and Laser Engineering,2014,43(3):985-990.(in Chinese)
[4]	[4] 王志社,杨风暴,纪利娥,等.基于聚类分割和形态学的可见光与SAR图像配准[J].光学学报,2014,34(2):0215002-1-7. WANG Z X,YANG F B,JI L E,et al.. Optical and SAR image registration based on cluster segmentation[J]. Acta Optica Sinica,2014,34(2):0215002-1-7.(in Chinese)
[5]	[5] 闫钧华,杭谊青,许俊峰,等.基于CUDA的高分辨率数字视频图像配准快速实现[J].仪器仪表学报,2014,35(2):380-386. YANG J H,HANG Y Q,XU J F,et al.. Quick realization of CUDA-based registration of high-resolution digital video images[J]. Chinese J. Scientific Instrument,2014,35(2):380-386.(in Chinese)
[6]	[6] LOWE D G. Distinctive image features from scale-invariant Key-points[J]. International J. Computer Vision,2004,60(2):91-110.
[7]	[7] PANCHAL P M,PANCHAL S R,SHAH S K. A comparison of SIFT and SURF[J]. Computer and Communication Engineering,2013,1(2):323-327.
[8]	[8] BAY H,TUYTELAARS T,VAN G L. SURF:speeded up robust features[J]. Computer Vision and Image Understanding,2008,110(3):346-359.
[9]	[9] 王健博,朱明.基于字典描述向量的实时图像配准[J].光学精密工程,2014,22(6):1613-1621. WANG J B,ZHU M. Real time image registration based on dictionary feature descriptor[J]. Opt. Precision Eng.,2014,22(6):1613-1621.(in Chinese)
[10]	[10] 张锐娟,张建奇,杨翠.基于SURF的图像配准方法研究[J].红外与工程,2009,38(1):160-165. ZHANG R J,ZHANG J Q,YANG C. Image registration approach based on SURF[J]. Infrared and Laser Engineering,2009,38(1):160-165.(in Chinese)
[11]	[11] MUJA M,LOWE D G. Scalable nearest neighbor algorithms for high dimensional data[J]. IEEE,2014,36(11):2227-2240.
[12]	[12] GUANG J S,XIANG Y X,YA P D. SIFT feature Point matching based on Improved RANSAC algorithm[C]. 2013 Fifth International Conference on Intelligent Human-Machine Systems and Cybernetics(IHMSC2013),Hangzhou,2013:474-477.