The Pavement Crack Detection System Based On Cloud Platform

At present,the traditional highway detection includes manual detection and multifunctional intelligent road detection vehicle detection.However,manual detection is laborious and inaccurate,multifunctional intelligent road detection vehicles are too expensive and have a single detection method.With the continuous development of the Internet,all walks of life are following the wave of the Internet,in order to achieve the industry's "change lanes overtaking." the combination with the Internet has important practical significance for the traditional road detection system.This paper mainly completed the development of crack detection software for vehicle camera,based on the software,a web-based service platform with the function of browsing and detecting road damage was developed,and the requirements of the web service end of the service program service were constructed.At the same time,the algorithm of crack extraction and damage statistics is designed,the algorithm includes image correction,non-scaling change filtering,image graying,image enhancement,image segmentation,and an interference object removal algorithm based on area method and proportion method according to the morphological characteristics of the crack.In the post-processing of crack,an improved version of the regional growth algorithm was implemented based on the study of the traditional regional growth algorithm to adapt to the remedy of the false elimination of micro cracks in this paper.Finally,a calculation method of crack damage parameters based on UK Scanner grid method and projection method is proposed to analyze the damage statistics and damage assessment of pavement.This paper first studies the main technology under the.NET platform,analyzesed its working principle and advantages and disadvantages,combines the functional requirements with its own functional requirements,and finally applies the latest two technologies into the system architecture of the web page.The front to ASP.NET MVVC framework based on Entity system performance,the back-end Framework framework for building a data access layer based on web by embedding Baidu map JSP,the Baidu road street map in the cloud upload and processing,data acquisition and data storage in the cloud damaged.Then,based on the socket network programming and multithreading technology to construct a simple service program,the main function of the image processing algorithm of service program with pictures,network communication protocol header processing,curlib library map resource requests based on Web,and the receiving end of the image data storage,processing and sending functions,service deployment in the program Ali cloud cloud server ESC,together with the deployment and web,database,image database.PC terminal software is based on MFC development and Halcon language is developed for algorithm development.The main modules include video processing module,GPS navigation module and data management.by the system test and algorithm verification,the results show that the system runs smoothly,the software operation is simple,the error of the road feature extraction algorithm and the repeating frame removal algorithm is acceptable,the communication between the web site and the server is stable.