Study On Camera Chain Vision Measurement Method Of Railway Subgrade Surface Settlement

Railway plays a critical role in the social and economic growth, while the relevant linear structure. And monitoring its surface settlement for the safety of railway operation possesses extraordinary significance. And non-contact visual measurement is the most favorable substitute artificial embankment settlement deformation observation.Firstly, camera chain visual detection method is proposed by ana-lysing the present status of the surface settlement detection for deformation, and introducing the vision measurement technology into the surface of the railway roadbed settlement tests for deformation, as well as using collimating characteristics of the laser. The method of coordinate system and its relevant transformation relationship is working and the measuring precision of this method is analysis-ed through simulation and experiment. By analyzing the accumulative relationship of the multi-stage transmission station’s measuring error, and the non-effectiveness of the transfer station’s positon modification to its subsequent testing result, this kind of method is handy and available in the railway subsidence deformation detection.Secondly, on the basis of traditional laser spot center orientation, to develop deep analysis about the image characteristics of the light spot to detect the edge of its flare and eliminate the ghost image by using gray mathematical morphology. Based on the analysis of by using the Gauss curve fitting method to calculate the flare’s sub-pixel centroid and flare’s gray distribution characteristics, the way about calculating flare’s sub-pixel centroid is proposed, according to the center of gravity’s gray curve distribution fitting method. Also, various flare’s impact on the positioning accuracy is tested through the simulation analyses, and two kinds of algorithms are certificated to be available by the actual experiment in the settlement of the visual inspection system.Then, the camera temperature drift method in the settlement optical inspection system is constituted by oversimplification the temperature drift model of the Handel’s. The effect of temperature change on the camera parameter needs to be measured through the given temperature experiment. The camera parameter temperature compensation model is developed by using that simplified model and system identification methods, and the effectiveness of this proposed model is also verified by related experiments. On this basis, the influence of temperature variation on the accuracy of the settlement measurement system is ana-lysed through the designed experiment.Also, for purpose of making breakthrough under the condition of being limited by the relay series in the chain of settlement deformation camera visual transmission measurement, this dissertation gives one kind of data optimization method that puts forward the double back and robustness is improved by adapting data optimization method based on the compensation method due to taking full advantage of this chain’s type of construction. And the settlement deformation of transmission network camera vision measurement and data optimization method based on the compensation method is advanced via deep expansion. And through the simulation analysis, the more of the inherent constraints can be used and the more recognizable image is to be submitted through improving the effect of the adjustment optimization method.Finally, the design realizes and proceeds the railway roadbed surface subsidence deformation camera chain and vision measurement system, the actual certification experiment and railway on-line monitoring tests. Through these actual experiment and railway on-line monitoring test, the accuracy and cumulative error of the measured settlement deformation monitoring is conforming to the requirement of the relevant procedures. This effectiveness and practicability of the measuring method and the measuring system also get tested and verified.