基于图像分析的桥梁混凝土结构施工裂缝检测技术

Bridge Concrete Structure Construction Crack Detection Technology Based on Image Analysis

由于在进行桥梁混凝土结构施工裂缝检测的过程中没有对裂缝图像进行灰度化、增强、阈值分割处理,致使检测的精确率、召回率和正确率较低,检测效果不佳,为此提出基于图像分析的桥梁混凝土结构施工裂缝检测方法。该方法首先通过双目立体视觉模型采集桥梁混凝土结构施工裂缝图像;其次,将采集的裂缝图像灰度化,并增强裂缝图像效果,对其进行阈值分割,提取其裂缝特征信息;最后,通过阈值方法结合投影计算裂缝的长、宽、面积等,完成对桥梁混凝土结构施工裂缝的检测。实验结果表明,所提方法的裂缝检测精确率、召回率和正确率均较高。

Due to the absence of graying,enhancement and threshold segmentation in the process of bridge concrete structure construction crack detection,the detection accuracy,recall and correctness are low,and the detection effect is not good.Therefore,a bridge concrete structure construction crack detection method based on image analysis is proposed.Firstly,the construction crack image of bridge concrete structure is collected by binocular stereo vision model.Then,the collected crack images are grayscaled,the crack images are enhanced,and the threshold segmentation is performed to extract the characteristic information of the cracks.Finally,the length,width,and area of the cracks are calculated using a threshold method combined with projection,completing the detection of construction cracks in bridge concrete structures.The experimental results show that the proposed method has a high precision rate,recall rate,and accuracy in crack detection.