A Preliminary Study On Non-Metric Digital Camera In Bridge Linear Recognition

With the rapid development of economy in China,the requirement for transportation is increasingly high.Besides,there are majority of hilly and mountain areas.In order to satisfy its demands,the number of bridge is also increasing year after year.While the deformation of bridge has a direct influence on the operation state,generally speaking,which can be measured by surveyor 's level,total station instrument and dial indicator.In the bridge monitoring,the traditional monitoring methods exist difficult to detect outside and the cost is expensive,which are more and more difficult to meet the needs of today's testing.With the rapid development of ordinary non-metric camera technology,the camera's resolution is constantly improved.It is of great practical significance to apply it to bridge monitoring.Therefore,through the study of the non-metric camera in the bridge monitoring,its work of the non-metric camera and the accuracy in the actual monitoring is improved,which has great economic value.Adopting traditional testing method can only check some important deformed location,while it is unable to obtain the whole deformation state.This paper uses the Canon EOS 5D? Mark camera as a tool for data acquisition,which not only combines the knowledge of close-range photogrammetry and image processing to apply it to linear recognition,but also Combines with the experiment to demonstrate its feasibility,The main research achievements as follows:(1)The plane calibration method is used to calibrate the camera to obtain the internal parameters.And then the coordinates of the points to be measured are obtained by the correlation method of the close range photogrammetry.Finally,the purpose of linear recognition can be realized.(2)Image conversion,enhancement and edge monitoring technology are used to deal with the image,remove the edge of the impurity extraction cable edge information,and then get the linear curve.And it can be studied through the use of different types of artificial markers,shooting distance,the size of artificial signs,shooting angle and the environment of the cable.Finally,we get a method to improve the accuracy of cable recognition.(3)Combined with the close range photogrammetry and image processing of the non-metric camera,the advantages of close range photogrammetry and image processing are fully utilized,so it can be applied to the monitoring of long-span bridges.