Research On Terrain Enhance Modeling With Linear Road’s Vector Data And Terrain Data In Plain And Hilly Area

The surface of the earth where we lived in is a complex and huge system, including not only the undulating terrain but also the complex and interrelated effects between to-pography and features on it. With the development of the three-dimensional visualization application, the realism of virtual geographic environments become increasingly demand-ing, how to display high precision terrain and features vividly have become the key of Vir-tual Geographic Environments Simulation. As an important typical feature, in existed three-dimensional data visualization system, the terrain of the linear road is descripted so simple that the visual result is far away from the real world, and that can’t meet the de-mand of people. For now, In some systems, like Google Earth, Virtual Earth or Baidu map, they simply map two-dimensional texture which contains terrain and features to DEM to simulate earth surface, that would reduce the quality of elevation data that along the road. It leads the road up and down irregularly, and three-dimensional performance of road dis-torted badly. This is not accord with people’s knowledge, and we can’t use the method to simulate the surface morphology of artificial structures such as road. Meanwhile, with the development of the earth observation technology such as photogrammetry via aircraft or satellite, remote sensing and GPS, varieties of geospatial data like DEM and high-precision imaging produces, these data have not been fully used. What’s more, in terms of terrain modeling, relying solely on High precision of data or manually modeling approach failed to meet a wide range and high level of detail, and many other needs at the time. So there are many researchers try to modeling terrain and features by integrate different data.At present, due to the lack of relevant theories and methods, large-scale integration of vector and raster data could not satisfy the increasing requirement. Some scholars re-searched the modeling of integrate vector road data and raster terrain data, but their meth-ods are poor in accuracy, the speed and the costs is also unacceptable. To improve the ex-pression of terrain in the plains and hilly areas effects, this paper presents a method for single-lineared road vector data and raster terrain data fusion modeling assisted by knowledge rules. The model reduction of road profiles and sections using DEM data as the underlying surface of high at the time of design, vector data of having been built highway as the road centerline, and the design and construction practices of roads at all levels as a secondary rules knowledge. The model is to reproduce the highway design and construc- tion processes based on all kinds of norms, standards and terrain data originally, and create the three-dimensional shape of the existed pavement, slope, drainage ditches and berms. This paper studies launched in the dimensions of building the knowledge rules pool of roads in the plains and hilly areas, optimizing and restoring of road’s cross-section and highway profile. Those are as follows:1) Classification and sorting the rule knowledge of roads in plains and hilly areaAccording to the ranking and type of the road, search specification and standards such as Road design specification People’s Republic of China (JTG D20-2006) and get the knowledge rules of longitudinal section and cross section. The knowledge rules of longitu-dinal section include maximum and minimum length of slope, maximum and minimum length of longitudinal slope, and radius of vertical curve. The knowledge rules of cross section include the width of roadbed, central strip, road shoulder and lane; the number of lane, and information of gutter, berm, side ditch and road humps.2) Optimize and restore of road’s cross sectionRoad’s cross section views is the slope graphic of the normal direction of the center line of road. It is surrounded by cross section design line and cross section ground line. Road’s cross section views is composed of lane, central strip, road shoulder, side slope, side ditch, drainage ditch, berm and so on. In this study, the cross section information of the experimental road consists of satellite image and result of fieldwork. Based on the re-sult of fieldwork and knowledge rules information, restore the road’s cross section.3) Optimize and restore of road’s longitudinal sectionAccording to the speed limit of different ranking roads in plain and hilly area, search relative design specification and standards and build a reasonable mathematic model to op-timize and restore of road’s longitudinal section. Then solve the model with AGA(adaptive genetic algorithm), analysis the characteristic of the algorithm, and finally achieve the op-timizing and restoring of road’s longitudinal section.In this paper, with a period of Nanjing airport highway as an experimental section the proposed method will be implemented systematically, and analyzed the error of results. The experimental results showed that the method proposed in this paper can get a road model data with fine visual effect. At the same time, its fine expression along the terrain of the road has been greatly improved in terms of elevation error compared with the DEM.