| 3D laser scanning technology which is also known as the "real copying technology", different from traditional measurement technology of single point data acquisition, this technology has many advantages, including high efficiency, abundant information, high precision, less influence of the external environment and non-contact,so it can obtain the high precision integral model. In recent years, 3D laser scanning technology is widely applied in the field of road reconnaissance design. By using this technology in survey of road reconstruction and extension can carry out the road survey of all-weather, measure with no interference, improve the work efficiency and reduce labor intensity. Based on the process of road construction and application of the technology of equipment type selection, data processing, feature extraction, this thesis mainly studies the following contents:(1) Search for suitable 3d laser scanning equipment selection and measurement method which are suitable for demand of road reconstruction and expansion.Study the principle of 3d laser scanner and its classification, analyze the sources of the measurement error and the corresponding measures to lower the error. According to the requirement for survey speed and accuracy and combined with the actual conditions, choose phase type Z + F Imager 5010 C scanner. Make the measurement scheme which is suitable for the demand of road reconstruction and expansion. This measurement scheme is combined of two ways, the features of the first pattern are 50 m station spacing, spherical target registration and high resolution, quality of normal mode. The features of the other pattern are 100 m station spacing, plane target registration, medium resolution and high quality model. Through the precision analysis, the maximum error of the splicing is 3.3 mm, which can meet the demand of road reconstruction and expansion.(2) Study the key algorithm of point cloud data processing which is applied to road reconstruction and expansion.Road reconstruction point cloud data processing mainly includes the point cloud registration, point cloud denoising, streamline compression, encapsulation and data mining, etc. This paper uses Geomagic Studio software, realizing point cloud registration by using the spherical target and plane target combination, realizing point cloud denoising by using artificial decision method, realizing point cloud simplification through combination of rule of distance and the rule of curvature. After encapsulation of point cloud, work on earthwork volume calculation method in the process of road construction by using the finite element integral principle of design, by using CASS and Rhinoceros software to realize the generation of DTM, introducing the related design parameters, generate the arrangement plan of road reconstruction and expansion(3) Research on feature extraction algorithm based on point cloud information wayIn order to do further research the application of three-dimensional laser scanning in the process of the reconstruction and expansion, select part of the completion of road to study the feature extraction algorithm. Using station spacing 30 ~ 50 m, target plane and spherical target comprehensive measurement scheme, obtain high precision model, design the wafer cutting algorithm of the point cloud, extract road geometrical characteristic parameters, the difference between the road plane, vertical and cross-sectional geometry parameters and the original design parameter is within 5 cm, only this way can meet the demand of road completion.This paper has formed a preliminary scheme of 3 d laser measurement and point cloud data processing algorithm in the process of survey and road reconstruction characteristic parameters extraction.It has certain reference value for further study of 3 d laser technology in this field. |