Research On The Method Of Measuring Bridge Deflection Based On Visual Image

As the lifeblood of my country's economy,road traffic construction plays an important role in the development of the country,and the scale and number of traffic projects are increasing day by day.my country's transportation construction is gradually transitioning from large-scale construction to equal emphasis on construction and maintenance,so maintenance and inspection at the operational stage will gradually become the focus of work.Bridge deflection monitoring is an important indicator of bridge safety evaluation.Most of the traditional displacement deflection measurement equipment is contact type,which depends on manpower to a certain extent and has high cost.There are limitations in actual engineering.With the rise of machine vision,image measurement methods with low cost,high precision,and non-contact advantages have been used in more and more fields.This article aims to study the method of measuring bridge deflection based on visual images.The specific research contents are as follows:(1)By analyzing the basic theories and methods related to digital image processing,the digital image is preprocessed,including the grayscale of the digital image,the binarization of the digital image,the edge detection of mathematical morphology,and the further denoising of the digital image,Digitize the image.(2)In order to improve the calculation speed of bridge measurement,a method of processing ROI for bridge video images is proposed.Then by comparing the three feature extraction(LBP feature,SIFT and ORB)of the image,the ORB feature matching that matches the actual working conditions is selected.Finally,the image correlation matching particle swarm algorithm is used for image correlation search.In order to further improve the bridge deflection measurement accuracy,the sub-cubic displacement search method of bicubic interpolation is used to achieve fast and high precision measurement of bridge deflection.(3)The imaging space theory of digital images realizes the conversion of space coordinates and camera coordinates.The Zhang's checkerboard method is used to perform calibration experiments on non-measurement cameras to obtain the internal and external parameters of the camera and the conversion coefficient between the image and the actual distance.Based on the previous research on image feature extraction and image feature matching,a bridge deflection extraction method based on visual images is proposed.Theexperiments of simulating the static displacement and dynamic displacement of the bridge prove that the method in this paper can be used in the actual measurement of bridge deflection and meet the accuracy requirements of bridge measurement.(4)Based on the research content of this article,a bridge deflection measurement and monitoring system based on visual images was preliminarily designed.Compared with the traditional contact bridge deflection measurement,the algorithm proposed in this paper realizes the non-contact bridge deflection measurement based on machine vision,which improves the accuracy,saves the engineering cost,and achieves the effects of real-time monitoring measurement and early warning.Compared with the existing image measurement algorithms,the image key area processing method,improved particle swarm optimization algorithm and sub-pixel search algorithm mentioned in this paper can improve the calculation speed of image processing while achieving deflection measurement.The image measurement method proposed in this paper achieves low-cost,real-time and high-precision deflection measurement of bridges,but this method fails to solve the impact of external factors such as camera installation position and environment on camera vibration in practical applications,and improves the accuracy of measurement results.Need further improvement.