摘要
基于数字图像处理技术,对采集的建筑物裂缝图片进行预处理,检测并识别裂缝,进而提取相关的裂缝特征参数。通过中值滤波方法消除灰度图像中的噪点,并采用最大熵值法对图像阈值分割,得到二值图像。经预处理后,利用边缘检测算法识别并提取图像裂缝边缘,编程定性判别裂缝走向及其形状特征,获取裂缝长度、最大宽度、平均宽度等特征参数。该方法可为建筑物结构安全性能的评定以及施工验收提供精确有效的数据基础。
Based on the digital image processing technology, the collected building crack images are preprocessed to detect and identify cracks, and then relevant crack characteristic parameters are extracted. The median filtering method is used to eliminate noise in gray images, and the maximum entropy method is used to segment the image threshold to obtain a binary image. After preprocessing, the Canny edge detection algorithm is used to identify and extract the edge of the image crack, and the program qualitatively identifies the crack trend and its shape features, and obtains the feature parameters such as crack length, maximum width and average width. This method will provide accurate and effective data basis for the evaluation of building structure safety performance and construction acceptance.
作者
周亚群
黄梦崖
邓林
ZHOU Yaqun;HUANG Mengya;DENG Lin(School of Mathematics and Statistics, Huangshan Univ., Anhui Huangshan 245041, China;School of Civil Engi. and Architecture, Huangshan Univ., Anhui Huangshan 245041,China)
出处
《低温建筑技术》
2018年第12期4-7,共4页
Low Temperature Architecture Technology
基金
安徽省教育厅自然科学研究项目(KJHS2018B02)
安徽省大学生创新创业项目(20170375033)
关键词
图像预处理
裂缝检测与提取
裂缝宽度
image preprocessing
crack detection and extraction
crack width
