ObjectLand supports three different geocoding methods. Choosing of one of them is determined by the task which should be solved by geocoding, by precision with which created features should be located on the map and by the data available for geocoding.
Geocoding by coordinates requires the presence of values of coordinates X and Y in two fields of geocoded table for its execution. When performing geocoding for every table record a point feature with indicated coordinates is created.
When performing geocoding by coordinates it is also possible to perform transformation of coordinates of the features being created. It is useful in the case if coordinates stored in the table are specified in coordinate system different from the one of the present map.
Geocoding by coordinates is the simplest geocoding method which at the same time allows the user to locate created features on the map with more precision.
Geocoding by features is based on binding of point features being created to the position of existing features of some type. Such geocoding method is used for example in the case when the geocoded table contains a list of addresses of objects (shops, service stations, etc.) required for the user.
Determining position of the feature by its address is the task which every person has to solve in practice. To solve it, it is enough to have a detailed map of the city where all buildings are located and their addresses are given. Then the position of the object searched for can be determined with precision to within a building.
In accordance with this analogy, apart from the source geocoded table and geocoded type it is required to indicate two more GDB components participating in the operation to perform geocoding by features in ObjectLand:
searching type – type of features to which position on the map features being created will be bound (for example, the type of features “Building” can be used as a searching type);
searching table – the table records of which are linked with features of searching type and which at the same time contain data allowing the user to associate a record of the searching table with each record of geocoded table. Thus, it is possible to find the corresponding feature of searching type for every record of geocoded table through searching table and use this feature for determining position of the feature created on the map.
If the searching type is “Buildings”, the table “Building Certificates” which records contain addresses of the corresponding buildings can be used as a searching table.
In comparison with geocoding by coordinates applying geocoding by features gives less precision but does not require explicit indication of coordinates of the features being created.
Geocoding by address is based on approximate calculation of position of the feature being created with respect to the specified line feature.
If we continue the analogy with using a city map, one can imagine the map on which addresses of particular buildings are not indicated, but maximum and minimum values of addresses of buildings in each block are specified for every street. If we suppose that buildings in some block have numbers from 100 to 120, the feature which is situated in building number 118 should be located close to the end of the block.
To perform geocoding by address in ObjectLand it is necessary to indicate two more GDB components participating in the operation apart from source geocoded table and geocoded feature type:
searching type – line type of features with respect to which position of created features on the map will be calculated. Feature type “Street segments”, for example, which contains segments of streets corresponding to separate blocks can be used as a searching type;
searching table of segments – the table records of which are firstly linked with features of searching type and secondly contain data allowing the user to associate a record of the searching table with each record of geocoded table and calculate the position of the feature being created.
If the searching type is “Street segments”, the table “Segments” records of which contain the street code and the range of addresses of buildings of the corresponding segment can be used as a searching table.
Precision of determining coordinates of features in geocoding by address depends on how much sizes of features corresponding to different addresses differ from one another. If building number 120 occupies half of the whole block in the given example, position of building number 118 will not be calculated very precisely.
For any geocoding method described above there is an additional possibility to set links between features being created and table records. The geocoded table itself as well as other GDB tables can be used as linked tables. In the last case the data from geocoded table is used for determining the record of another table to be linked with the feature being created.
 |  |
