Rapid Recognition of Rock Mass Fractures in Tunnel Faces
LENG Biao1,2, ZHANG Yi3, YANG Hui1,2, HOU Gaopeng1,2
1. School of Civil Engineering，Southwest Jiaotong University，Chengdu 610031，China; 2. Key Laboratory of Transportation Tunnel Engineering，Ministry of Education，Southwest Jiaotong University，Chengdu 610031，China; 3. China Railway Second Engineering Group Co.，Ltd.，Chengdu 610031，China
Abstract：Tunnel faces contain much geological information，which，if fully extracted and analyzed，will help to evaluate the geological state of tunnel engineering and guide tunnel design and construction. In this paper，rock mass crack detection，extraction and grouping algorithms of the tunnel face are studied on the basis of the tunnel face image. First，the image segmentation algorithm for face rock mass cracks is analyzed using digital image processing technology. According to the segmentation results，the discontinuous boundaries are connected and the short boundaries are filtered through image thinning and boundary fitting，separation，merging and filtering to form a relatively complete recognition result of rock mass boundary lines. Then，the apparent dip angle of rock mass boundary line is calculated，and the boundary lines with similar apparent dips are merged into a group to achieve automatic grouping. This method is applied to real tunnel face images to test its effectiveness. Results show that the proposed method can basically realize automatic extraction and grouping of the rock mass cracks in the funnel face. For rock masses with obvious cracks，the algorithm can extract the cracks with an error rate of more than 10% and implement automatic grouping with an error rate of no and 5%. This method improves automation degree of rock mass analysis of the tunnel face，and can be used for geological sketching map，providing references for the surrounding rock classification of tunnel faces.
李术才,刘洪亮,李利平,等. 基于数码图像的掌子面岩体结构量化表征方法及工程应用[J]. 岩石力学与工程学报,2017,36(1): 1-9LI Shucai, LIU Hongliang, LI Liping, et al. A quantitative method for rock structure at working faces of tunnels based on digital images and its application[J]. Chinese Journal of Rock Mechanics and Engineering, 2017, 36(1): 1-9
[2]
LEU S S, CHANG S L. Digital image processing based approach for tunnel excavation faces[J]. Automation in Construction, 2005, 14: 750-765
[3]
周春霖，朱合华，李晓军. 新奥法施工隧道掌子面红外照相及图像处理[J]. 岩石力学与工程学报，2008，27（增刊1）：3166-3172.ZHOU Chunlin，ZHU Hehua，LI Xiaojun. Application of infrared photography and image processing to tunnel construction with new austrian tunneling method[J]. Chinese Journal of Rock Mechanics and Engineering，2008，27（S1）：3166-3172.
[4]
李鹏云,赵科,陈孜迪,等. 基于图像处理的隧道掌子面地质构造信息提取研究[J]. 土木建筑工程信息技术,2017,9(6): 67-72LI Pengyun, ZHAO Ke, CHEN Zidi, et al. Research on extraction of geological structure information of tunnel excavation faces based on image processing[J]. Journal of Information Technologyin Civil Engineering and Architecture, 2017, 9(6): 67-72
[5]
范留明,李宁. 基于数码摄影技术的岩体裂隙测量方法初探[J]. 岩石力学与工程学报,2005,24(5): 792-797FAN Liuming, LI Ning. Study on rck mass joint measurement based on digital photogrammetry[J]. Chinese Journal of Rock Mechanics and Engineering, 2005, 24(5): 792-797
[6]
罗佳,刘大刚. 围岩结构面发育程度参数的图像处理技术研究[J]. 计算机科学与工程,2013,35(4): 75-80LUO Jia, LIU Dagang. Research on image processing technology of developmental degree parameters of surrounding rock structure[J]. Computer Science and Engineering, 2013, 35(4): 75-80
[7]
叶英,王梦恕. 隧道掌子面地质信息数字编录识别技术研究[J]. 北京交通大学学报,2007,31(1): 59-62YE Ying, WANG Mengshu. Research on digital catalogue identification technology of geological information in tunnel face[J]. Journal of Beijing Jiaotong University, 2007, 31(1): 59-62
[8]
胡刚,金乾坤. 岩体天然裂隙计算机图像处理技术研究[J]. 有色金属(矿山部分),2004,56(5): 39-40HU Gang, JIN Qiankun. Research on computer image processing technology of natural fracture in rock mass[J]. Nonferrous Metals (Mineral Part), 2004, 56(5): 39-40
[9]
李弼程，彭天强，彭波，等. 智能图像处理技术[M]. 北京：电子工业出版社，2004：153.
[10]
CANNY J. A computational approach to edge detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1986, 8(6): 679-698
