Research And Application Of Defect Detection Method For Prefabricated Building Precast Concrete

In recent years,prefabricated buildings have developed rapidly on the road of transformation and upgrading of the domestic construction industry.Quality control is the key link of prefabricated construction projects and the basic guarantee for construction applications.As the core components of prefabricated buildings,the quality of precast concrete directly determines whether the construction can be successfully implemented.The missing of reserved holes and surface cracks of precast concrete are common defects.If they are not found and treated in time,they may cause great economic losses to enterprises.At present,enterprises mainly use the traditional manual method for defect detection,which is not only cumbersome and inefficient,but also difficult to ensure the quality of detection.Therefore,how to ensure the quality of precast concrete and improve the efficiency of defect detection through automation and intelligent technical means is particularly important.At present,the defect detection technology based on computer vision has been widely used in the field of industrial defect detection.It has the characteristics of good flexibility,high accuracy and strong robustness,which is very suitable for defect detection applications in industrial scenarios.Therefore,this thesis uses computer vision technology to design corresponding detection methods for the defects of reserved hole tooling and surface cracks in the production process of precast concrete,and applies them to the prefabricated building defect detection management system,so as to improve detection efficiency,reduce production costs,and ensure the quality of finished products.This thesis mainly focuses on three aspects: the detection method for the missing tooling of reserved holes in precast concrete,the identification and measurement method of surface cracks in precast concrete,and the defect detection management system for prefabricated building precast concrete.The main work of this thesis is as follows:(1)Aiming at the defect of missing tooling of reserved holes,a detection device and method for missing tooling of reserved holes based on multi-target tracking are designed.Among them,the detection device and detection method work together.The detection device is responsible for automatically collecting the data of the precast concrete to be detected,and then handing it to the detection method for real-time processing,so as to judge whether the precast concrete is qualified.In order to improve the accuracy and stability of multi-target tracking algorithm,the generalized intersection ratio is used to construct the cost matrix.In addition,a tool location and matching method is proposed to accurately find the missing tool.In this thesis,the target detection algorithm model is trained and evaluated on the self-constructed reserved hole tooling detection data set,and the superiority of the improved algorithm and the validity of the detection method are verified on the reserved hole tooling tracking data set.(2)Aiming at the surface crack defects of precast concrete,a crack identification and measurement method of prefabricated building precast concrete is proposed.The method combines deep learning technology with traditional digital image processing technology to realize accurate identification and extraction of cracks,and uses the pixel calibration technology based on QR code to get the actual physical size of Cracks,so as to complete the measurement of crack size.In this thesis,the target detection algorithm model and image classification algorithm model in the proposed method are trained and evaluated respectively on the precast concrete crack detection data set and precast concrete crack recognition data set constructed by ourselves,and the effectiveness of the proposed method is verified by crack measurement experiments.(3)Aiming at the problems of difficult management of quality inspection data,asynchrony of multi-party information,and difficulty in tracing quality problems,a defect detection management system for prefabricated building precast concrete was designed and implemented.The system can not only detect the two defects of the missing of reserved hole tooling and surface cracks of precast concrete,but also provides flexible integrated management functions of the system and comprehensive quality inspection data management functions,which can effectively improve the efficiency of quality inspection related personnel.In this thesis,computer vision technology is applied to the defect detection of precast concrete of prefabricated buildings.Aiming at two defects of precast concrete and the quality inspection work problems of enterprises,the defect detection method and system of precast concrete are designed and implemented to assist quality inspection staff to carry out reasonable and efficient defect detection.