摘要
针对数字全景钻孔摄像系统获取的实测图像,提出了一种基于特征函数的钻孔图像结构面识别方法。该方法根据井壁结构面在钻孔图像中呈正弦曲线状的特征,采用正弦函数作为特征函数来匹配识别结构面。通过改变特征函数参数来统计分析结构面及附近像素点灰度梯度的匹配情况,然后筛选出一条匹配最优的正弦曲线作为该结构面的特征曲线,从而实现了钻孔图像结构面的识别。该方法省略了传统方法中人工选取结构面特征点和特征点曲线拟合的过程,其方法稳定可靠、准确率高,为钻孔图像结构面的自动识别与工程应用提供了一种实际可行的有效途径。
For the borehole image obtained by digital panoramic borchole camera system, a method for recognizing structural plane is described based on characteristic function. According to the sinusoidal shaped characteristics of borehole side-wall discontinuities in the image, the method adopts standard sine function as characteristic function. The recognition of borehole structural plane is realized by changing the function parameters to match the nearby pixel grey gradient distribution and by choosing the best matching sine curve as the characteristic curve of structure plane. The method ignores the traditional process of manual selecting characteristic points and fitting characteristic curve. It provides a practical and effective way for the automatic recognition of borehole structural planes in engineering application with good stability and high accuracy.
作者
王川婴
邹先坚
韩增强
汪进超
王益腾
WANG Chuan-ying ZOU Xian-jian HAN Zeng-qiang WANG Jin-chao WANG Yi-teng(State Key Laboratory of Geomechanics and Geotechnical Engineering, Institute of Rock and Soil Mechanics, Chinese Academy of Sciences, Wuhan, Hubei 430071, China School of Electronic Information, Wuhan University, Wuhan, Hubei 430072, China)
出处
《岩土力学》
EI
CAS
CSCD
北大核心
2017年第10期3062-3066,共5页
Rock and Soil Mechanics
基金
国家自然科学基金(No.41402278
No.41372317)~~
关键词
钻孔图像
结构面
特征函数
迭代匹配
borehole image
structural plane
characteristic function
iterative matching
