Research On Detection Technique Of Facilities Along Roads Based On Stereoscopic Photography

With the rapid development of China’s highway construction, the detection of highwayroadbed, road tunnel, bridge, culvert structures and the facilities is paid more and moreattention.How to detect the damage project of the highway facilities quickly, objectively andaccurately, which has great significance to our highway management and maintenance.This paper makes a detailed analysis on the present situation of the highway facilitiesdetection technology and equipment at home and abroad and puts forward the highwayfacilities detection method of lateral convergent photogrammetry based on stereophotogrammetry. The method can choose camera structure layout according to detectionneeds, avoid the too small field and the image occlusion. According to the research on theroad facilities detection technology based on the stereoscopic photography, the paperfocuses on the following aspects:Firstly, based on the basic principle of stereo photogrammetry, the paper analyses thestereo vision system structure and adopts lateral convergent photogrammetry to detect thehighway facilities. Designs the general structure of the road facilities detection systembased on the stereoscopic photography combined with the requirements of the highwayfacilities detection system and introduces the selected method of the main hardwareequipments.Secondly, the paper deeply researches two key technologies of the highway facilitiesdetection system which are camera calibration and image processing. Camera calibrationadopts direct linear transform algorithm to calculate the camera inside and outsideparameters, through the camera inside and outside parameters and image space coordinatecomputs object space coordinate. Image processing adopts manual annotation method tomake image processing of the facilities, and calculates geometry parameters according tothe type of the facilities damaged project.Thirdly, in accordance with the requirements of the facilities detection system and userrequirements, the paper designs the collecting software, processing software, analysingsoftware and other assisted softwares required by detection system.Finally, the platform is set up and indoor experiments are made to verify the methods of camera calibration and image processing. In the outdoor environment, the parametersobtained by camera calibration and the datas of the highway facilities are collected,processed and analysed.The above research verifies that the road facilities detection system based on thestereoscopic photography can make a fast, objective and accurate detection on the roadfacilities.