Support for Create 3D Computer Graphics Images in GIS Systems Using Externals Renders Algorithm A Review of Robust Estimation Methods Applied in Surveying

Browsing the possibilities posed by GIS and CAD programs in the modeling analysis results we can find many similarities and differences. Features arise from aims for which these types of programs have been developed. Goal of GIS software is to do different types of spatial analysis. A secondary aim is to do visualization of the results generated. A secondary aim is to doing visualization of the results generated. CAD software is taking data presentation proposed as an overriding goal, however, analysis tools, topology and relationship to the database are limited to special overlays to the main program or did not exist. Knowing how to do the main task of particular programs we can say that the CAD software is better suited for the visualization of data than GIS applications. We will attempt to analyze possibility of these two types of software with regard to ArcGIS Version 10 and CAD graphics software Cinema4D R13. For visualization in the program ArcGIS we can use two 3D environments: ArcGlobe and ArcScene. They are designed for different uses. After starting ArcGlobe application, we can see the globe, suggests to us, that all analysis and visualization will involve a large area of land. The data can be presented on a sphere Earth. ArcScene is designed for analysis and visualization covering a small area of land. Data are presented on the local reference plane. Differences also include data that we can use for analysis and visualization. In ArcGlobe, all data can be displayed in the specified simplification, depending on the magnification. This eliminates the problem of overloading the computer memory, we do not have to resign from less important data available. In ArcScene, all available data are automatically stored in the computer memory, we have quick access to them. The downside of

Browsing the possibilities posed by GIS and CAD programs in the modeling analysis results we can find many similarities and differences.Features arise from aims for which these types of programs have been developed.Goal of GIS software is to do different types of spatial analysis.A secondary aim is to do visualization of the results generated.A secondary aim is to doing visualization of the results generated.CAD software is taking data presentation proposed as an overriding goal, however, analysis tools, topology and relationship to the database are limited to special overlays to the main program or did not exist.Knowing how to do the main task of particular programs we can say that the CAD software is better suited for the visualization of data than GIS applications.
We will attempt to analyze possibility of these two types of software with regard to ArcGIS Version 10 and CAD graphics software Cinema4D R13.
For visualization in the program ArcGIS we can use two 3D environments: ArcGlobe and ArcScene.They are designed for different uses.After starting ArcGlobe application, we can see the globe, suggests to us, that all analysis and visualization will involve a large area of land.The data can be presented on a sphere Earth.ArcScene is designed for analysis and visualization covering a small area of land.Data are presented on the local reference plane.Differences also include data that we can use for analysis and visualization.In ArcGlobe, all data can be displayed in the specified simplification, depending on the magnification.This eliminates the problem of overloading the computer memory, we do not have to resign from less important data available.In ArcScene, all available data are automatically stored in the computer memory, we have quick access to them.The downside of

Introduction
Despite the continuous improvement of survey methods and advances made in survey equipment technology, the elimination of outliers still remains an issue today.When performing an adjustment one often assumes a very simple probability distribution of errors, such as a normal distribution.In classical statistics the correctness of the results relies on the assumption, that the chosen errors distribution model is strictly true.This is, in fact, often not the case, as the large errors occur considerably more often than the normal distribution would suggest.Even the high-quality samples analysed in astronomical research, containing several thousands of measurements each, do not follow the normal probability distribution.Deviations from the model may occur due to e.g.blunders in measuring, incorrect point numbering, errors made during data copying etc. [12].
Although there exists a wide range of literature concerned with gross errors detection and elimination, this surveying problem is still being discussed.There are many so-called methods robust against the in uence of gross errors, which can generally be divided into two groups.
The  rst group includes methods based on the criteria of so-called robust estimation.These methods minimise the in uence of the outlying observations on the  nal result of the computations by modifying of the observation weights.
The second of them consists of methods where results, obtained by the least squares adjustment are analysed with the use of statistical tests.In these methods an identi ed outlier is removed from the dataset.If multiple outliers occur, the iterative process of least squares adjustment is conducted and followed by tests.The observations suspected of gross errors are discarded from the dataset [1].A few commonly used methods of these groups are presented below.
this is dependence of the maximum amount of data used to the quantities computer memory.Exceeding the limit of memory deteriorates the smooth operation of program.Many other differences between ArcGlobe and ArcScene software presented in the Table 1.
After comparing the options above we can conclude that the best tool for visualizing data on a small range, for example, the analysis of place of the buildings and the analysis of hydrographic, is a software ArcScene.In the settings display quality graphics in ArcGlobe and ArcScene is available only poorly developed illumination option responsible for the lighting.It includes elements such as the angle of incidence of solar radiation, altitude and position of the simulated sun, and contrast.Poor number of options theory gives us to understand that the generated images are deviate significantly from the required photo realism.
ArcScene software will be analyzed as a dedicated tool for visualizing data on a small range for example, the analysis of the location the building and the analysis of hydrographic.In ArcScene, layers can be textured by any images using the symbolization.Texture on one layer may differ between themselves after a proper classification, or use of unique symbols for the layer arising from the database.Lighting the whole scene consists in dimming or brightening the textures depending on the angle and height of the sun.The big downside is the lack of shadows cast by the illuminated object on the other layers, such as for example, houses do not cast shadows on the ground, what makes the visualized scene the artificial.An additional problem is the lack of antialising settings, i.e. smoothing the edges of objects.This minus may be removed by setting the appropriate option in the video card drivers.This will cause partial smoothing out all the edges of the escalator.Another downsides of the ArcScene are limited opportunities of modeling objects.A simple 3D model can be created from 2D polygon by extrude.Extrusion turns polygons into blocks.transparent elements and textures on surfaces.Without it, the visualization of certain types of analyzes cannot be properly understood.
For enhancing the att ractiveness of our scenes, we can use a number of simulations of atmospheric physics eff ects, which are polluted air, fog, smoke, water vapor and clouds.These eff ects allow us to get bett er control over the photorealism scene, emphasizing the problem and simulation of phenomena occurring on the visualized area.Examples are air pollution in the city, moving fog in mountain ranges, fi re and smoke in areas aff ected by the fi res.
Using multiple eff ects it also has its drawbacks.For one of them we can include rendering time.Some scenes require generation of a few hours.Another downside is that the excess of diff erent eff ects can obscure the main issue presented in the image of the analysis performed.Therefore, all eff ects and glamourizing the scene should be chosen carefully in order not to unnecessarily lengthen the rendering time and not to overshadow, but highlight the problem presented.In Figure 3 we can see an example of the algorithm renderer, contained in the CAD software, in the process of visualization objects holiday village.The picture was created using the GIS spatial data about the location of each element, and objects (trees, texture) were added at a later stage in the CAD software Cinema 4D.Time image rendering lasted about 30 minutes.

Cooperation ArcGIS Software with the Program Cinema4D
By doing the analysis of various types, to a lot of software, we are not forced to use for everything only one software.There is no question that the program will ArcScene bett er suited to do analyzes, prepare data for visualization, while the rendering algorithm application Cinema4D -to generate an image presenting the results of analyzes.

Fig. 3 .
Fig. 3. Visualize the location and placement buildings designed holiday resort

Table 1 .
Selected differences between ArcGlobe and ArcScene applications