[1]杨婷婷,管昉立,徐爱俊*.基于Graph Cut算法的多株立木轮廓提取方法[J].南京林业大学学报(自然科学版),2018,42(06):091-98.[doi:10.3969/ j.issn.1000-2006.201804012]
 YANG Tingting,GUAN Fangli,XU Aijun*.Multiple trees contour extraction method based on Graph Cut algorithm[J].Journal of Nanjing Forestry University(Natural Science Edition),2018,42(06):091-98.[doi:10.3969/ j.issn.1000-2006.201804012]
点击复制

基于Graph Cut算法的多株立木轮廓提取方法
分享到:

《南京林业大学学报(自然科学版)》[ISSN:1000-2006/CN:32-1161/S]

卷:
42
期数:
2018年06期
页码:
091-98
栏目:
研究论文
出版日期:
2018-12-15

文章信息/Info

Title:
Multiple trees contour extraction method based on Graph Cut algorithm
作者:
杨婷婷1管昉立2徐爱俊1*
(1. 浙江农林大学信息工程学院,浙江省林业智能监测与信息技术研究重点实验室,林业感知技术与智能装备国家林业与草原局重点实验室,浙江 杭州 311300; 2. 武汉大学,测绘遥感信息工程国家重点实验室,湖北 武汉 430079)
Author(s):
YANG Tingting1GUAN Fangli2XU Aijun1*
(1.School of Information Engineering, Zhejiang Provincial Key Laboratory of Forestry Intelligent Monitoring and Information Technology, Key Laboratory of State Forestry and Grassland Administration on Forestry Sensing Technology and Intelligent Equipment, Zhejiang A & F University, Hangzhou 311300, China; 2. State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, China)
关键词:
Graph Cut算法 轮廓提取 改进Canny算子 图像分割
Keywords:
Graph Cut algorithm contour extraction improved Canny operator image segmentation
分类号:
S781; TP391.41
DOI:
10.3969/ j.issn.1000-2006.201804012
文献标志码:
A
摘要:
【目的】在复杂的自然环境下进行目标立木轮廓提取时,容易受遮挡物影响,导致立木图像分割效果不理想。笔者提出一种基于Graph Cut算法的多株立木轮廓提取方法,可实现单张相片中多目标立木界线分割。【方法】首先通过彩色直方图均衡化实现RGB颜色空间下各个通道的图像细节增强,利用Graph Cut算法构造s-t网络图,将图像分割问题转化为能量函数最小化问题,并标记图像前背景像素实现单张相片中多株立木图像初分割; 然后将单张相片中的每株立木分割图像二值化,利用形态学腐蚀膨胀运算处理图像达到填充、去噪、平滑等目的; 在此基础上,利用改进型Canny算子边缘检测方法,用双边滤波代替高斯滤波增强边界信息得到每株立木轮廓; 最后,根据立木相对坐标不变性,利用几何重组方法实现目标立木特征表达并判断其拓扑关系,最终得到每株目标立木轮廓提取结果。【结果】为了验证该方法的有效性,本研究对自然环境下采集到的立木图像进行试验。结果表明,该方法能够在不同光照条件的复杂背景下,有效分割出每株立木轮廓,平均误分率为5.62%,假阳性率为4.49%,假阴性率为4.33%,均优于常用的OTSU分割算法(41.40%、26.73%、10.99%)、K-means聚类算法(49.97%、35.02%、11.92%)和基于C-V模型水平集法(28.43%、20.53%、13.38%)。【结论】复杂的自然环境下,利用基于人工交互的Graph Cut算法可有效分割出每株立木轮廓界,研究结果可为立木可视化重建、特征提取等提供参考。
Abstract:
【Objective】 Because of the complexity of the natural environment, current tree contour extraction results are not satisfactory. This paper presents a method to extract a contour of multiple trees based on a Graph Cut algorithm to realize the boundary segmentation of multi-target trees in a single photo. 【Method】 First, this method enhances image details of each channel under RGB color space captured in the experiment by color histogram equalization. The graph of the s-t network is constructed using a Graph Cut algorithm to look for min-cut, and the image segmentation problem is transformed into the minimization of the energy function by marking the foreground and background pixels to achieve a single photo of many trees with preliminary image segmentation. Then, adaptive thresholding of gray-scale transformation is applied to the multiple-segmentation images to realize binarization of the images and morphological corrosion expansion, and the opening and closing operation processing of the binary images is used to achieve the filling, denoising, and smoothing of trees. On the basis of this morphological process, combined with the improved Canny operator edge detection technology, bilateral filtering is used instead of Gaussian filtering to enhance the boundary information to obtain a preliminary tree contour. Finally, according to the geometric position invariance of the photo’s trees, we use the geometric reconstruction method to express the features of target trees and judge their topological relationships. If there are topological relationship errors, we iterate the Graph Cut algorithm and geometry reassembly method again to obtain a better target tree extraction result. 【Result】 In order to validate the effectiveness of this method experimentally, we collected tree images in a natural environment. The results showed that this method can effectively separate the contour of every tree under different lighting conditions. The average error rate(Af)was 5.62%, the false positive rate(RFP)was 4.49%, and the false negative rate(RFN)was 4.33%, which was better than those obtained by the traditional OTSU segmentation algorithm(41.40%, 26.73% and 10.99%, respectively), the K-means clustering algorithm(49.97%, 35.02% and 11.92%), and the C-V Plane Models(28.43%, 20.53% and 13.38% ). 【Conclusion】 In a complex natural environment, a Graph Cut algorithm based on human interaction can effectively separate vertical boundaries. The results provide a reference for the visualization and reconstruction of trees and feature extraction.

参考文献/References:


[1] 周克瑜,汪云珍,李记,等. 基于Android平台的测树系统研究与实现[J]. 南京林业大学学报(自然科学版),2016,40(4): 95-100. DOI: 10.3969/j.issn.1000-2006.2016.04.015.
ZHOU K Y,WANG Y Z,LI J,et al. A study of tree measurement systems based on Android platform[J]. Journal of Nanjing Forestry University(Natural Sciences Edition),2016,40(4): 95-100.
[2] 王建利,李婷,王典,等. 基于光学三角形法与图像处理的立木胸径测量方法[J]. 农业机械学报,2013,44(7): 241-245.DOI: 10.6041/j.issn.1000-1298.2013.07.042.
WANG J L,LI T, WANG D,et al. Measuring algorithm for tree’s diameter at breast height based on optical triangular method and image processing[J]. Transactions of the Chinese Society for Agricultural Machinery,2013,44(7): 241-245.
[3] 沈亚峰,王歆晖,巩宇涵,等. 基于智能手机图像分析的树木胸径测量研究[J]. 江苏林业科技,2017,44(1): 28-33.DOI: 10.3969/j.issn.1001-7380.2017.01.006
SHEN Y F,WANG X H,GONG Y H,et al. Study on measurement of tree diameter size based on smartphone imageanalysis[J]. Journal of Jiangsu Forestry Science & Technology,2017,44(1): 28-33.
[4] 赵茂程,郑加强,林小静,等. 基于分形理论的树木图像分割方法[J]. 农业机械学报,2004,35(2): 72-75.
ZHAO M C,ZHENG J Q,LIN X J,et al. Tree image segmentation method based on the fractional dimension[J]. Transactions of the Chinese Society for Agricultural Machinery,2004,35(2): 72-75.
[5] 张军国,冯文钊,胡春鹤,等. 无人机航拍林业虫害图像分割复合梯度分水岭算法[J]. 农业工程学报,2017,33(14): 93-99.DOI: 10.11975/j.issn.1002-6819.2017.14.013.
ZHANG J G,FENG W Z,HU C H,et al. Image segmentation method for forestry unmanned aerialvehicle pest monitoring based on composite gradien twatershed algorithm[J]. Transactions of the Chinese Society of Agricultural Enginee-ring,2017,33(14):93-99.
[6] 王晓松,黄心渊,付慧. 复杂背景下的树木图像提取[J]. 北京林业大学学报,2010,32(3):197-203.DOI: 10.13332/j.1000-1522.2010.03.019.
WANG X S, HUANG X Y, FU H.Study surveys on tree image extraction in a complex background[J]. Journal of Beijing Forestry University,2010,32(3): 197-203.
[7] 刁智华,王欢,宋寅卯,等. 复杂背景下棉花病叶害螨图像分割方法[J]. 农业工程学报,2013,29(5): 147-152.DOI: 1002-6819(2013)-05-0147-06.
DIAO Z H,WANG H,SONG Y M, et al. Segmentation method for cotton mite disease image under complex background[J]. Transactions of the Chinese Society of Agricultural Engineering,2013,29(5): 147-152.
[8] 李冠林,马占鸿,黄冲,等. 基于K-means硬聚类算法的葡萄病害彩色图像分割方法[J]. 农业工程学报,2010,26(2): 32-37.DOI: 10.11975/j.issn.1002-6819.2010.02.005.
LI G L,MA Z H,HUANG C, et al. Segmentation of color images of grape diseases using K-means clustering algorithm[J]. Transactions of the Chinese Society of Agricultural Engineering,2010,26(2):32-37.
[9] 杨信廷,孙文娟,李明,等. 基于K均值聚类和开闭交替滤波的黄瓜叶片水滴荧光图像分割[J]. 农业工程学报,2016,32(17): 136-143.DOI: 10.11975/j.issn.1002-6819.2016.17.019.
YANG X T,SUN W J,LI M,et al. Water droplets fluorescence image segmentation of cucumber leaves based on K-means clustering with opening and closing alternately filtering[J]. Transactions of the Chinese Society of Agricultural Enginee-ring,2016,32(17): 136-143.
[10] 贺付亮,郭永彩,高潮,等. 基于视觉显著性和脉冲耦合神经网络的成熟桑葚图像分割[J]. 农业工程学报,2017,33(6): 148-155. DOI: 10.11975/j.issn.1002-6819.2017.06.019.
HE F L,GUO Y C,GAO C, et al. Image segmentation of ripe mulberries based on visual saliency and pulse coupled neural network[J]. Transactions of the Chinese Society of Agricultural Engineering,2017,33(6): 148-155.
[11] 徐黎明,吕继东. 基于同态滤波和K均值聚类算法的杨梅图像分割[J]. 农业工程学报,2015,31(14): 202-208.DOI: 10.11975/j.issn.1002-6819.2015.14.028.
XU L M,Lü J D. Bayberry image segmentation based on homomorphic filtering and K-means clustering algorithm[J]. Transactions of the Chinese Society of Agricultural Engineering,2015,31(14): 202-208.
[12] BARGOTI S, UNDERWOOD J P. Image segmentation for fruit detection and yield estimation in apple orchards[J]. Journal of Field Robotics, 2016, 34(6):14-20.
[13] WANG C, TANG Y, ZOU X, et al. A robust fruit image segmentation algorithm against varying illumination for vision system of fruit harvesting robot[J]. Optik-International Journal for Light and Electron Optics, 2017(131):626-631.DOI: 10.1016 /j.ijleo.2016.11.177.
[14] 宋怀波,张传栋,潘景朋,等. 基于凸壳的重叠苹果目标分割与重建算法[J]. 农业工程学报,2013,29(3): 163-168.DOI: 10.11975/j.issn.1002-6819.2013.03.022.
SONG H B,ZHANG C D,PAN J P,et al. Segmentation and reconstruction of overl appedapple images based on convexhull[J]. Transactions of the Chinese Society of Agricultural Engineering,2013,29(3): 163-168.
[15] 王新忠,韩旭,毛罕平. 基于吊蔓绳的温室番茄主茎秆视觉识别(英文)[J]. 农业工程学报,2012,28(21): 135-141,294. DOI: 10.3969/j.issn.1002-6819.2012.21.019.
WANG X Z,HAN X,MAO H P. Vision-based detection of tomato main stem in greenhouse with red rope[J]. Transactions of the Chinese Society of Agricultural Engineering,2012,28(21): 135-141,294.
[16] 程玉柱,陈勇,张浩. 基于MMC与CV模型的苗期玉米图像分割算法[J]. 农业机械学报,2013,44(11): 266-270. DOI: 10.6041/ j. issn.1000-1298. 2013.11.045.
CHENG Y Z,CHEN Y,ZHANG H. Color image segmentation algorithm of corn based on MMC and CV model[J]. Transactions of the Chinese Society for Agricultural Machinery,2013,44(11): 266-270.
[17] 冯静静,张晓丽,刘会玲. 基于灰度梯度图像分割的单木树冠提取研究[J]. 北京林业大学学报,2017,39(3): 16-23. DOI:10.13332/j.1000-1522.20160373.
FENG J J,ZHANG X L,LIU H L. Single tree crown extraction based on gray gradient image segmentation[J]. Journal of Beijing Forestry University,2017,39(3): 16-23.
[18] BOLLANDSAS O M,NAESSET E,SOLBERG S. Single tree segmentation using airborne laser scanner data in a structurally heterogeneous spruce forest[J]. Photogrammetric Engineering & Remote Sensing,2006,72(12):1369-1378.
[19] KWATRA V, ESSAI, TURK G, et al. Graphcut Textures: image and video synthesis using graph[J]. Acm Transactions on Graphics, 2003, 22(3): 277-286.
[20] KUO J W, MAMOU J, WANG Y, et al. Segmentation of 3-D high-frequency ultrasound images of Human lymph nodes using Graph Cut with energy functional adapted to local intensity distribution[J]. IEEE Trans Ultrason Ferroelectr Freq Control,2017,64(10): 1514-1525. DOI:10.1109/TUFFC.2017.2737948.
[21] CHEN X,UDUPA K,BAGCI U,et al. Medical image segmentation by combining graph cuts and oriented active appearance models[J]. IEEE Trans Image Process,2012,21(4): 2035-2046. DOI:10.1109/TIP.2012.2186306.
[22] ZHOU H,ZHENG J,WEI L. Texture aware image segmentation using graph cuts and active contours[J]. Pattern Recognition,2013,46(6): 1719-1733. DOI:10.1016/j.patcog.2012.12.005.
[23]CIESIELSKI K C,UDUPA J K,FALC?O A X,et al. Fuzzy connec-tedness image segmentation in Graph Cut formulation: A linear-time algorithm and a comparative analysis[J]. Journal of Mathematical Imaging and Vision,2012,44(3): 375-398. DOI:10.1007/s10851-012-0333-3.
[24] YIN S,ZHAO X,WANG W,et al. Efficient multilevel image segmentation through fuzzy entropy maximization and graph cut optimization[J]. Pattern Recognition,2014,47(9): 2894-2907. DOI:10.1016/j.patcog.2014.03.009.
[25] CIESIELSKI KC,MIRANDA PA,FALC?O AX,et al. Joint graph cut and relative fuzzy connectedness image segmentation algorithm[J]. Med Image Anal,2013,17(8): 1046-1057. DOI:10.1016/j.media.2013.06.006.
[26] 苏金玲,王朝晖. 基于Graph Cut和超像素的自然场景显著对象分割方法[J]. 苏州大学学报(自然科学版),2012,28(2): 27-33.
SU J L, WANG Z H. An image segmentation method based on graph cut and super pixed in nature scene[J]. Journal of Suzhou University(Natural Science Edition), 2012,28(2): 27-33.
[27] 韩守东,赵勇,陶文兵,等. 基于高斯超像素的快速Graph Cuts图像分割方法[J]. 自动化学报,2011,37(1): 11-20.DOI: 10.3724/SP.J.1004.2011.00011.
HAN S D,ZHAO Y,TAO W B, et al. Gaussian super-pixel based fast image segmentation using graph cuts[J]. Acta Automatica Sinica,2011,37(1): 11-20.
[28] 韦轶群,杨杰. 基于自适应Graph Cuts的自动股骨头分割[J]. 上海交通大学学报,2009,43(3): 465-469.DOI:10.16183/j.cnki.jsjtu.2009.03.027.
WEI Y Q,YANG J. Automatic segmentation of femoral heads based on adaptive Graph Cuts[J]. Journal of Shanghai Jiaotong University,2009,43(3): 465-469.
[29] 刘毅,黄兵,孙怀江,等. 利用视觉显著性与图割的图像分割算法[J]. 计算机辅助设计与图形学学报,2013,25(3): 402-409.
LIU Y,HUANG B,SUN H J, et al. Image segmentation based on visual saliency and graph cuts[J]. Journal of Computer-Aided Design & Computer Graphics,2013,25(3): 402-409.
[30] 孙龙清,李玥,邹远炳,等. 基于改进Graph Cut算法的生猪图像分割方法[J]. 农业工程学报,2017,33(16): 196-202.DOI: 10.11975/j.issn.1002-6819.2017.16.026.
SUN L Q,LI Y,ZOU Y B,et al. Pig image segmentation method based on improved Graph Cut algorithm[J]. Transactions of the CSAE,2017,33(16): 196-202.
[31] 周良芬,何建农. 基于GrabCut改进的图像分割算法[J]. 计算机应用,2013,33(1): 49-52.
ZHOU L F,HE J N. Improved image segmentation algorithm based on Grab Cut[J]. Journal of Computer Applications,2013,33(1): 49-52.
[32] 董茜,颜凯,孙婷婷,等. 基于SLIC超像素的Grab Cut算法改进[J]. 华中科技大学学报(自然科学版),2016,44(S1): 43-47,66. DOI:10.13245/j.hust.16S109.
DONG Q,YAN K,SUN T T, et al. Improvement of grab cut algorithm based on SLIC superpixel algorithm[J]. J Huazhong Univ of Sci & Tech(Natural Science Edition), 2016,44(S1): 43-47,66.
[33] 陈鑫,何中市,李英豪. 一种新的基于SLICO改进的GrabCut彩色图像分割算法[J]. 计算机应用研究,2015,32(10): 3191-3195. DOI: 10.3969/j. issn.1001-3695.2015.10.073.
CHEN X,HE Z S,LI Y H. Improved color image segmentation of GrabCut algorithm based on SLICO[J]. Application Research of Computers,2015,32(10): 3191-3195.
[34] 周志宇,刘迎春,张建新. 基于自适应Canny算子的柑橘边缘检测[J]. 农业工程学报,2008,24(3): 21-24.
ZHOU Z Y,LIU Y C,ZHANG J X. Orange edge detection based on adaptive Canny operator [J]. Transactions of the CSAE,2008,24(3): 21-24.
[35] 段红燕,邵豪,张淑珍,等. 一种基于Canny算子的图像边缘检测改进算法[J]. 上海交通大学学报,2016,50(12): 1861-1865. DOI:10.16183/j.cnki.jsjtu.2016.12.009.
DUAN H Y,SHAO H,ZHANG S Z,et al. An improved algorithm for image edge detection based on Canny operator[J]. Journal of Shanghai Jiaotong University,2016, 50(12): 1861-1865.
[36] 贺强,晏立. 基于LOG和Canny算子的边缘检测算法[J]. 计算机工程,2011,37(3): 210-212.
HE Q,YAN L. Algorithm of edge detection based on LOG and canny operator[J]. Computer Engineering,2011,37(3): 210-212.

备注/Memo

备注/Memo:
收稿日期:2018-04-09 修回日期:2018-08-07
基金项目:国家自然科学基金项目(31670641); 浙江省科技重点研发计划(2018C02013)
第一作者:杨婷婷(917251944@qq.com),硕士。*通信作者:徐爱俊(xuaj1976@163.com),教授,博士。
更新日期/Last Update: 2018-11-30