This document explains the method to geocode using street address data in NT Visualiser
Geocoding and Labelling using street address information
NT Visualiser is available across NTG and is an application which marries the functionality and imagery found in Google Earth with spatial information and imagery managed and owned by NTG. For more information (including installation and log in) go to NT Visualiser - Information and Installation Manual
Prior to Geocoding, it is important that you collect accurate, current and complete street address information. For the purpose of this demonstration, a text file called "AddressText.txt" has been created (see NT Visualiser - Geocoding street addresses example) and will be used. The text file has delibrately included misspellings, missing data and incomplete street addresses to demonstrate what happens when incomplete or inaccurate data is used.
Steps
- Open NT Visualiser and log in
- Go to File > Open and navigate to where "LGA with XY.txt" is saved on your workstation
- Change “Files of Type” to Generic Text (*.txt *.csv), the text file should appear
- Select AddresstTest.txt
- Click Open
The Data Import Wizard
appears (Figure 1)
Figure 1 - The Data Import Wizard
As we are opening a text file, the following parameters need to be se
For Field Type - Tick "Delimited"
For Delimited - Tick "Treat consecutive delimiters as one"
For Delimited - Tick "Tab"
Looking at the Preview of the data, it is evident that already, with the omission of a street type, all the data has been moved one place to the right. This will cause problems later on (as we will see)
- Click Next
The Select Latitude/Longitude Field Window appears(Figure 2)
Figure 2 - Select Latitude/Longitude Field Window
In this window,
Tick "This dataset does not contain latitiude/longitude information, but street addresses
- Click Next
- The next Window ( Figure 3) asks if there are Street Address attributes in your text file. Choose one of the options depending on your data structure. For this example, Select "The Dataset contains one address field"
- Address field value = full address ( corresponding column in the text file)
- Click Back
Figure 3 - Definition of Street Addresses Window
The Specify Field Types (optional) window appears ( Figure 4). All values can remain as is. Click Back
Figure 4 - Specify Field Type (optional) Window
- The Definition of Street Addresses appears again (Figure 3). Don't change any settings
- Click Finish
If all street number, street name, street type and suburb values are complete and correct, the following error should not appear
Although the error does not tell you what exactly is the problem (or its cause), the street name "Chewing" is spelt incorrectly. The correct spelling is "Chewings". To fix, cancel this message ( click the red X box), open the text file, correct the mistake, save and then repeat all steps outlined before this point.
The following message appears if all the above configuration has been completed correctly. Click Yes to apply a style template
- In the Style Template Options Window, tick "create new template" and Click OK. If an existing style template exists, it will be listed under "Compatible Templates" List. You can use these templates and/or edit them.
The Style Template Settings
Window (Figure 5) appears next. Here, you can configure the colour and allocate a desired icon for placemarkers
We would like to label the street number for each parcel that we are geocoding. To enable labelling set the value of "Set name field" to street_no
(or the column which has the information you wish to be labelled)
Figure 5 - The Style Template Settings Window
Review the Color and Icon tabs and configure these settings as desired. An Icon represents a placemarker. Each geocoded address will be represented by a placemarker at or near the parcel boundaries or at the centre of the parcel.
Click OK once you are satisfied with the settings
Save your Template. By default it will save in D:\Document and Settings\<your user name>\Application Data\Google\Google Earth\My Style Templates\ as a .kst file extension
Once saved, NT Visualiser will refresh and
The map window should re centre and displayed the imported placemarkers with the street number as a label
A new KML called "Address.txt" will appear under Places>Temporary Places in the left hand side
Figure 6 shows the result of importing and geocoding "AddressText.txt" into NT Visualiser.
Figure 6 - Results of importing geocoding street address and labelling street numbers
The Addresses defined in the text file are geocoded and matched to address data provided through a service from Google. The currency of this data provided by Google is uncertain. There may be some differences in data currency between addresses provided by Google and those found in Government web mapping systems or ILIS ( the later being more current).
Geocoding - Things to look out for
From the example above, four addresses geocoded correctly (7,8,9 and 5). Three did not
The following table shows the data in the full_address column of AddressText.txt ( data used to Geocode) and whether it Geocoded or not. Reasons have been provided to highlight possible erorrs that can cause Geocoding to not work
full_address | Geocoded |
---|---|
7 Kirk Place Gillen | Yes - Geocoded correctly and represents the correct parcel |
8 Nicker Crescent Gillen | Yes - Geocoded correctly and represents the correct parcel |
9 Ballingall Street The Gap | Yes - Geocoded correctly and represents the correct parcel |
5 GAP ROAD THE GAP | Yes - Geocoded correctly and represents the correct parcel |
9 Chewing Street East Side | No - "Chewing" misspelt so did not Geocode |
43 Cavenagh Street East Side | Yes - 43 does not exist so it has geocoded to the road centreline |
40 Cavenagh East Side | No - missing street type. Did not import at all. |
5/14 Smith Street Alice Springs | No - unit numbers cannot be Geocoded. Omit |
This table highlights the importance of ensuring that when you are collecting street address data, you must ensure that the street numbers exist, the street name and type are spelt correctly and exist and the suburb is actually a registered place name