全景图像镶嵌中累积误差的分析与减少方法研究

计算机工程与科学

全景图像镶嵌中累积误差的分析与减少方法研究

吴琼，李良福，汪之涛，肖樟树

（陕西师范大学计算机科学学院，陕西西安 710119）

收稿日期:2015-05-26 修回日期:2015-11-04 出版日期:2016-11-25 发布日期:2016-11-25
基金资助:
国家自然科学基金(61201434,61573232,61401263)；中央高校基本科研业务费专项资金(GK201503068)

An accumulated error analysis and a reduction

method for panoramic mosaic images

WU Qiong,LI Liangfu,WANG Zhitao,XIAO Zhangshu

（School of Computer Science,Shaanxi Normal University,Xi’an 710119,China）

Received:2015-05-26 Revised:2015-11-04 Online:2016-11-25 Published:2016-11-25

摘要/Abstract

摘要：

针对全景图像镶嵌中累积误差的问题，提出了一种基于多次柱面投影的变换基准图策略。首先，为了将累积误差有效地分散到各相邻图像之间，并减少柱面投影变换的计算量，算法利用相位相关和多分辨率分解法对柱面投影算法进行了改进；然后，利用改进变换基准图的分层匹配策略对图像序列进行镶嵌，避免了误差的单向累积；最后，运用拉普拉斯算子对镶嵌图像分级融合，消除了镶嵌缝隙。实验结果表明，该方法可以大大减弱镶嵌过程中累积误差的影响，提高图像的镶嵌精度。

关键词: 改进柱面投影, 全景图像镶嵌, 累积误差, 分散误差

Abstract:

In order to solve the problem of the accumulated error in panoramic mosaic images, we propose a reference image transformation strategy based on multiple cylindrical projections. Firstly, we improve the cylindrical projection algorithm by combining the phase correlation and the multiresolution decomposition to effectively distribute the accumulated error of the adjacent image and reduce the amount of computation in the cylindrical projection transformation. Secondly, we adopt the strategy of hierarchical matching based on reference image transformation to perform image sequence mosaic, which can avoid unidirectional error accumulation in the mosaic. Finally, we utilize the Laplacian operator to carry out step fusion, which generates highquality results and renders a seamless output panorama. Experimental results show that the proposed strategy can greatly reduce the impact of accumulated error in the process of mosaic and improve mosaic precision.

Key words: improved cylindrical projection, panoramic image mosaic, accumulated error, distribution error

吴琼，李良福，汪之涛，肖樟树. 全景图像镶嵌中累积误差的分析与减少方法研究[J]. 计算机工程与科学.

WU Qiong,LI Liangfu,WANG Zhitao,XIAO Zhangshu.

An accumulated error analysis and a reduction

method for panoramic mosaic images

[J]. Computer Engineering & Science.

参考文献

［1］Hui Fengming, Cheng Xiao, Liu Yan, et al. An improved landsat image mosaic of antarctica［J］.Science China（Earth Sciences），2013，01（8）:112．

［2］Ma J,Zhang C,Luo X.An effective approach for generating 360degree panorama views［C］∥Proc of IEEE International Conference on Management of Ecommerce & Egovernment,2011:290293．

［3］Li Pengjun,Li Jianzeng,Song Yao,et al.Error analysis of UAV remote sensing image mosaic based on simulation［J］.Journal of Computer Applications,2015,4（6）:116119.（in Chinese）

［4］Marzotto R,Fusiello A,Murino V.High resolution video mosaicing with global alignment［C］∥Proc of the 2004 IEEE Conference on Computer Vision and Pattern Recognition,2004:692698．

［5］Kim D W,Hong K S.Realtime mosaic using sequential graph［J］.Journal of Electronic Imaging,2006,15（2）:2329．

［6］Healy D M,Rohde G K.Fast global image registration using random projections［C］∥Proc of the 4th IEEE International Symposium on Biomedical Imaging,2007:476479．

［7］Tang Jin, Gu Shiwen, Cai Zixing. Theoretical analysis on image stitching error［J］.MiniMicro Systems，2003，24（12）:21852189.（in Chinese）

［8］Li Congli,Xue Mogen,Leng Xiaoyan,et al.Analysis and elimination on aerial recon sequential image stitching accumulative error［J］.Journal of Image and Graphics，2008，13（4）:814819.（in Chinese）

［9］Lu Heng,Li Yongshu,He Jing, et al. An automatic mosaic method in unmanned aerial vehicle images based on feature points［J］.Geography and GeoInformation Science,2010,26（5）:1619．(in Chinese)

［10］He Jing, Li Yongshu, Lu Heng. Experimental study on UAV image stitching error［J］.Laser & Optoelectronics Progress,2011,48（12）:6468．(in Chinese)

［11］Liu Qingyuan,Liu You,Zou L,et al.UAV’s image quick mosaic and its errors［J］.The Chinese Journal of Nonferrous Metals,2012,22（3）:889894．(in Chinese)

［12］Wu Yu,Yu Tao,Xie Donghai.Fast automatic stitching for images of unmanned aerial vehicle in emergency response［J］.Journal of ComputerAided Design & Computer Graphics,2013,25（3）:410416．(in Chinese)

［13］Xu Yuhu,Gong Wenquan,Su Ang,et al.Multiviewpoint image mosaicing algorithm for roughly planar scenes［J］.
Journal of National University of Defense Technology,
2014,

36（2）:148155．(in Chinese)

［14］Kuglin C,Hines D.The phase correlation image alignment method［C］∥Proc of IEEE Conference on Cybernetics and Society,1975:163165．

［15］Xia Y M,Liu Y X,Chen S L,et al.A rectangular shell element formulation with a new multiresolution analysis［J］.Acta Mechanica Solida Sinica,2014,06（9）:612615．

附中文参考文献：

［3］李鹏俊,李建增,宋瑶,等.基于仿真的无人机遥感图像拼接误差分析［J］.计算机应用,2015,4（6）:116119．

［7］唐琎,谷士文,蔡自兴.图象拼接误差的理论分析［J］.小型微型计算机系统，2003，24(12):21852189.

［8］李从利,薛模根,冷晓艳,等.空中侦察序列图像连续拼接的累积误差分析与消除［J］.中国图象图形学报，2008，13(4):814819．

［9］鲁恒, 李永树，何敬, 等. 一种基于特征点的无人机影像自动拼接方法［J］. 地理与地理信息科学,2010,26（5）:1619.

［10］何敬, 李永树，鲁恒. 无人机影像拼接误差实验研究［J］. 激光与光电子学进展,2011,48（12）:6468.

［11］刘庆元, 刘有, 邹磊, 等.无人机影像的快速拼接及其误差［J］. 中国有色金属学报,2012,22（3）:889894.

［12］吴俣, 余涛, 谢东海. 面向应急响应的无人机图像快速自动拼接［J］. 计算机辅助设计与图形学学报,2013,25（3）:410416.

［13］徐玉华，龚文全，苏昂，等.近似平面场景多视点图像拼接算法［J］.国防科技大学学报,2014,36（2）:148155．