高级检索

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于间隙度的无人机林地航拍图像序列拼接方法

张一茗 付慧

张一茗, 付慧. 基于间隙度的无人机林地航拍图像序列拼接方法[J]. 北京林业大学学报, 2017, 39(6): 107-115. doi: 10.13332/j.1000-1522.20170020
引用本文: 张一茗, 付慧. 基于间隙度的无人机林地航拍图像序列拼接方法[J]. 北京林业大学学报, 2017, 39(6): 107-115. doi: 10.13332/j.1000-1522.20170020
ZHANG Yi-ming, FU Hui. A method for woodland aerial image sequences mosaic based on lacunarity[J]. Journal of Beijing Forestry University, 2017, 39(6): 107-115. doi: 10.13332/j.1000-1522.20170020
Citation: ZHANG Yi-ming, FU Hui. A method for woodland aerial image sequences mosaic based on lacunarity[J]. Journal of Beijing Forestry University, 2017, 39(6): 107-115. doi: 10.13332/j.1000-1522.20170020

基于间隙度的无人机林地航拍图像序列拼接方法

doi: 10.13332/j.1000-1522.20170020
基金项目: 

教育部高等学校优秀青年教师研究基金项目 2015ZCQ-XX

详细信息
    作者简介:

    张一茗。主要研究方向:图像处理、计算机视觉。Email:amingmao0117@163.com  地址:100083 北京市海淀区清华东路35号北京林业大学信息学院

    责任作者:

    付慧, 博士, 副教授。主要研究方向:图像拼接、图像识别。Email:fuhuir@bjfu.edu.cn  地址: 同上

  • 中图分类号: S771.8

A method for woodland aerial image sequences mosaic based on lacunarity

  • 摘要: 无人机林地航拍图像具有的分辨率高、数据量大、边缘丰富的特点,造成了特征点配准中误匹配率的增加,因此本文提出了一种新的无人机林地航拍图像序列拼接方法。分形中的间隙度可用于描述图像区域纹理的粗糙程度,本文首先利用间隙度特征选取图像中局部图像块作为候选区域查找特征点,减少了待配准的特征点数量,提高了特征点配准正确率;其次,采用全局拼接技术变换图像,减少传统拼接中矩阵连乘产生的误差的积累和传播。实验中选取了不同拍摄高度的两组图像序列,将本方法与传统的全局SURF特征方法和降采样图像拼接方法进行了对比,结果显示本方法可以有效拼接图像,同时不会损失原图像的精度信息,并从视觉效果和均方根误差两个角度证明了本文方法优于其他两种方法。

     

  • 图  1  对整幅林地航拍图像进行特征配准时产生的误匹配的特征点示例

    4种不同颜色标记表示4对误匹配的特征点。

    Figure  1.  Wrong matching points with SURF matching

    4 different color marks represent 4 wrong matching points.

    图  2  局部图像块配准流程图

    Figure  2.  Flowchart of registration of local image block

    图  3  候选图像块示例

    Figure  3.  Candidate image blocks

    图  4  目标图像中候选图像块区域扩展

    block:图像块

    Figure  4.  Expanding the candidate image block in target image

    Image block

    图  5  全局图像变换示意图

    Figure  5.  A diagram of global image registration method

    图  6  实验1对比结果展示

    重叠区域高达90%, 图像被分为8 × 8块。

    Figure  6.  Comparison results of experiment 1

    Overlap area reaches 90% and the image is divided into 8 × 8 blocks.

    图  7  实验2对比结果展示

    重叠区域仅有60% ~ 70%, 图像被分为6 × 6块。

    Figure  7.  Comparison results of experiment 2

    Overlap area is only 60%-70%, and the image is divided into 6 × 6 blocks.

    表  1  第1组实验部分RMSE值比较

    Table  1.   Comparison of 1st image sequence sets

    序列Sequence LAC-SURF SURF 降采样Down-sample
    1 75.96 3 562.9 1 107.05
    2 50.33 3 177.6 1 932.01
    下载: 导出CSV

    表  2  第2组实验部分RMSE值比较

    Table  2.   Comparison of 2nd image sequence sets

    序列
    Sequence
    LAC-SURFSURF 降采样
    Down-sample
    1 468.39 2 338.4 3 502.53
    2 443.21 2 619.2 2 850.40
    下载: 导出CSV
  • [1] LIU J, CHEN Z Y, GUO R F. A mosaic method for aerial image sequence by R/C model[C]//Proceedings of International Conference on Computer Science and Software Engineering 2008. Washington D C: IEEE Computer Society Press, 2008: 58-61. https://www.researchgate.net/publication/224362282_A_Mosaic_Method_for_Aerial_Image_Sequence_by_RC_Model
    [2] BOTTERILL T, MILLS S, GREEN R. Real-time aerial image mosaicking[C]//Proceedings of International Conference of Image and Vision Computing New Zealand (IVCNZ) 2010 25th. Washington D C: IEEE Computer Society Press, 2010: 1-8. https://www.researchgate.net/publication/228963263_Real-time_Aerial_Image_Mosaicing
    [3] YIN X, MA N N.Study on the automatic unmanned aerial vehicle image mosaic algorithm[C]//Proceedings of International Conference on Intelligent Control and Information Processing (ICICIP) 2011 2nd. Washington D C: IEEE Computer Society Press, 2011: 624-628. https://www.researchgate.net/publication/241190378_Study_on_the_automatic_Unmanned_Aerial_Vehicle_Image_mosaic_algorithm
    [4] TRIGGS B, MCLAUCHLAN P F, FITZGIBBON A W, et al. Bundle adjustment: a modern synthesis[M]//TRIGGS B, ZISSERMAN A, SZELISKI R.Vision algorithms: theory and practice. Berlin:Springer, 1999: 298-372.
    [5] ULLMAN S. The interpretation of structure from motion[J].Journal of the Royal Society of London, 1979, 203: 221-226. http://d.old.wanfangdata.com.cn/OAPaper/oai_arXiv.org_1110.0424
    [6] 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. http://d.old.wanfangdata.com.cn/Periodical/gfkjdxxb201402024

    XU Y H, GONG W Q, SU A, et al. Multi-viewpoint image mosaicing algorithm for roughly planar scenes[J]. Journal of National University of Defense Technology, 2014, 36(2):148-155. http://d.old.wanfangdata.com.cn/Periodical/gfkjdxxb201402024
    [7] 谭郁松.自然图像处理理论和方法研究[D].长沙: 国防科技大学大学, 2004.

    TAN Y S. Research on natural images processing theory and method[D]. Changsha: journal of National University of Defense Technology, 2004.
    [8] MANDELBROT B B. The fractal geometry of nature[M].New York: Freeman, 1982. https://www.researchgate.net/publication/238344516_The_Fractal_Geometry_Of_Nature
    [9] QUAN Y, XU Y, SUN Y, et al. Lacunarity analysis on image patterns for texture classification[C] //Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Washington D C: IEEE Computer Society Press, 2007: 160-167. https://www.researchgate.net/publication/286594095_Lacunarity_Analysis_on_Image_Patterns_for_Texture_Classification
    [10] DU G, YEO T S. A novel lacunarity estimation method applied to SAR image segmentation[J]. IEEE Transactions on Geoscience & Remote Sensing, 2002, 40(12): 2687-2691. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=ba7e4cdefe8f7ade1b19abf5321988e4
    [11] MUJA M. Fast approximate nearest neighbors with nutomatic algorithm configuration[C]//Proceedings of International Conference on Computer Vision Theory and Applications(VISAPP). Washington D C: IEEE Computer Society Press, 2009: 331-340.
    [12] 李庆忠.大视场视频全景图快速生成方法[J].计算机工程, 2009, 35(22): 170-184. doi: 10.3969/j.issn.1000-3428.2009.22.058

    LL Q Z. Fast construction method for large field video view panorama[J]. Computer Engineering, 2009, 35(22): 170-184. doi: 10.3969/j.issn.1000-3428.2009.22.058
    [13] 王勃, 龚志辉, 顾广杰, 等.一种快速的无人机影像无缝拼接方法[J].测绘科学, 2012, 37(5): 23-26. http://d.old.wanfangdata.com.cn/Periodical/chkx201205007
  • 加载中
图(7) / 表(2)
计量
  • 文章访问数:  1315
  • HTML全文浏览量:  248
  • PDF下载量:  22
  • 被引次数: 0
出版历程
  • 收稿日期:  2017-01-18
  • 修回日期:  2017-03-13
  • 刊出日期:  2017-06-01

目录

    /

    返回文章
    返回