Researches On Extraction Of Slope Deformation Information Based On TLS

Slope engineering is an ancient and ever new problem in construction process.With the vigorous development of China's infrastructure,there has been a lot of slope problems in mining,irrigation,municipal,construction,roads,railways and other areas.Bulletin of china Land and Resources has showed that Landslides and collapses caused by slope failure often bring huge losses of life and property.Slope monitoring is an important part of the slope research,through the slope monitoring to determine the unstable slope sliding mode,sliding direction and speed,master the slope changes,and take prevention measures timely,try to avoid and reduce engineering and personnel losses.So slope monitoring is very important for safety of slope engineering.With the development of science and technology,all kinds of advanced testing instruments and equipment,monitoring methods and monitoring methods are constantly updated,and the level of the slope monitoring is continuously improved.In this paper,based on the principle of 3D laser scanning technology,the application of 3D laser scanning technology in the extraction of slope deformation information is studied.The main research contents are:(1)Introduced measuring principle and present status of application of 3D laser scanning technology,analyses the factors that influence 3D laser scanner point position precision,The point position error model of a three dimensional laser scanner is derived based on the measurement of the distance,angle and the influence of the spot.(2)In this paper,forfour-element algorithm requires at least three of the same name point and ICP registration accuracy is low,and an four parameter-ICP point cloud registration model is proposed.The four-parameter-ICP point cloud registration algorithm is to use two same name points with four parameters for initial registration,re-use ICP registration algorithm again.The registration model was tested by the scanning point cloud data,and the experimental results show that the four parameter-ICP point cloud registration model can reduce the registration error.Using a variety of methods to extract the coordinates of the center point of the plane target and the ball target,the results show that the fuzzy C-mean clustering method and the total least square method can obtain the high accuracy of the center point coordinates.(3)This paper presents a new method to determine the DMI by which precursory displacements can be detected.A core of the approach is the computation of point position error of the deformation orientation based on the achievement of actual point error ellipsoid.Firstly,the computation of point error ellipsoid using the error standard deviation of the angle and range as well as the influence of the laser spot;then,the calculation of point cloud error ellipsoid considering the relationship between the neighboring point error ellipsoids;and then,the computation of the actual point error ellipsoid and point position error;Finally,the determination of functional relationship between DMI and point position error.Determining the DMI of point cloud at any sampling interval,distance,and incidence angle greatly improves the reliability of deformation detection.(4)The three-dimensional laser scanning technique is used to carry out the scanning experiment of an artificial slope in Wuhan.Using thefour parameter-ICP model of point cloud registration on the slope point cloud data,extracting Point and regional deformation information of the slope.At the same time,the threshold value of the 3D laser scanner in the deformation monitoring is calculated,and the real deformation of the slope is obtained.The results show that the slope deformation information extracted by TLS(Terrestrial Laser Scanning)System is reliable.