基于计算机方法的混凝土识别与应用研究  

Research on Concrete Identification and Application Based on Computer Method

在线阅读下载全文

作  者:赵发宾[1] ZHAO Fabin(Science and Technology Department,Kunming Metallurgy College,Kunming 650033,China)

机构地区:[1]昆明冶金高等专科学校,云南昆明650033

出  处:《粉煤灰综合利用》2020年第3期29-32,共4页Fly Ash Comprehensive Utilization

摘  要:为提高混凝土裂缝识别检测的效率,建立卷积神经网络模型,并结合MATLAB对混凝土裂缝形态开展识别。当图像进行灰度处理后,通过理论公式得到了图片特征信息,分析出混凝土裂缝识别效率与二值化阈值的关系,建议二值化阈值取0.3~0.6之间;采用3种检测算子进行裂缝边缘检测,通过对比得到了各检测算子的适用范围:拉普拉斯算子更适用于灰度不同区域,索贝尔算子的识别效率与二值化阈值有关,边缘检测算子能够全面的反映出边缘裂缝情况。卷积神经网络图像识别方法能够为混凝土结构的安全状态评估提供理论和技术支持。In order to improve the efficiency of concrete crack identification and detection,a convolution neural network model was established,and the concrete crack morphology was identified with MATLAB.When the image was gray-scale processed,the image feature information was obtained by theoretical formula;the relationship between the recognition efficiency of concrete cracks and the binarization threshold was obtained by setting different binarization thresholds,and it was suggested that the binarization threshold should be between 0.3 and 0.6;three detection operators were used in crack edge detection,and the application scope of each detection operator was obtained by comparison:Laplacian operator is more suitable for different regions of gray scale.The recognition efficiency of Sobel operator is related to the binarization threshold,and edge detection operator can fully reflects the situation of edge cracks.The convolution neural network image recognition method can provide theoretical and technical support for the safety state assessment of concrete structures.

关 键 词:神经网络 混凝土 检测 识别 应用 

分 类 号:TP391.41[自动化与计算机技术—计算机应用技术] U456.3[自动化与计算机技术—计算机科学与技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象