隧道裂缝自动识别性能影响因素研究  

Study of factors affecting automatic identification performance of cracks in tunnel

作　　者：段中兴[1] 周建[1] 陈胜[1] 裴可[1] 

机构地区：[1]西安建筑科技大学信息与控制工程学院,陕西西安710055

出　　处：《传感器与微系统》2016年第10期60-62,66,共4页Transducer and Microsystem Technologies

基　　金：国家自然科学基金资助项目(50575168);陕西省教育厅自然科学专项基金资助项目(07JK281)

摘　　要：在室内环境下,利用混泥土试验模块,设计一种基于互补金属氧化物半导体(CMOS)传感器件的装置来进行隧道裂缝自动识别试验。将裂缝图像灰度值分布作为衡量自动识别的重要指标,分析有效像素、检测距离、光照强度等因素对自动检测性能的影响。模拟试验结果表明:有效像素和检测距离对图像分布的特征影响不大;有效像素增大,检测距离减小,相应的检测性能增加;光照强度不仅对裂缝图像灰度分布特征影响大,对自动检测性能也有显著的影响,光照强度过高或过低都会影响检测性能。In indoor environment,using concrete test module,design a device based on CMOS sensor to automatically identify cracks in tunnel test. Crack image gray value distribution is regarded as an important indicator for automatic recognition. Analyze effective pixels,testing effect of distance,light intensity and other factors on automatic detection performance. Simulation results show that the effective pixel and detection distance have little effect on features of image distribution. Effective pixels is increased,detecting distance is reduced,corresponding detecting performance is increased; light intensity not only has great impact on fracture image gray distribution feature,but also has a significant impact on automatic detection performance,too high or too low light intensity will affect detecting performance.

关 键 词：隧道裂缝 互补金属氧化物半导体传感器 自动识别 检测性能 

分 类 号：U451[建筑科学—桥梁与隧道工程]