摘要
为提高计算机对膨胀土裂隙图像的识别精度,针对膨胀土裂隙的特点,引入边缘检测技术对膨胀土的裂隙图像进行处理,从一阶微分算子到二阶微分算子进行膨胀土裂隙的试验。结果表明与其他算子比较,Canny算子在图像噪声的去除以及避免裂隙信息失真上具有较好效果。
In order to improve the accuracy of image recognition of expansive soil by computer, by using the edge detection technique dealing with the fissures image of expansive soil, the tests of expansive soil fissures are carried out from the first differential operators to the two order differential operators. The results show that compared with other operators, Canny operator has better effect in image noise removal and crack information distortion recognition.
作者
陈永青
方炫强
陈宗辉
孙政
王勉
Chen Yongqing;Fang Xuanqiang;Chen Zonghui;Sun Zheng;Wang Mian(Hunan Provincial Key Laboratory of Engineering Rheology, Central South University of Forestry and Technology, Changsha 410018, China;Hunan Provincial Communications Planning, Survey ＆ Design Institute, Changsha 410018, China)
出处
《湖南文理学院学报（自然科学版）》
CAS
2018年第2期58-62,共5页
Journal of Hunan University of Arts and Science(Science and Technology)
基金
湖南省重点学科建设项目(2013ZDXK006)
中南林业科技大学引进高层次人才科研启动基金(104–0094)
关键词
远程监测
膨胀土裂隙
图像识别
多技术融合
remote monitoring
fissure of expansive soil
image recognition
multi technology fusion
