KUBEL logo KUBEL logo    SECTION5
KUBEL logo KUBEL logo Home | Embedded / Signal Processing | Device control library | Visualization | Contacts/Links

round
Overview
Lightmapping
DIY maps

Creating Your Own Maps


Introduction

When planning a longer trip to yet unknown places, you might be left with some topological questions: how to get there, where are the next cities, where are the mountains. Often, you can not obtain detailed maps from local suppliers, even worse, some countries do not even provide maps for various reasons. So, one has to sometimes bite into the sour apple and make his/her own map, since the apple should not be to sour, it is not desirable to pay large amounts of money for raw data. Well, this is not necessary, as there are free sources, listed below.

Obtaining the data

The sources

Table 1. Data sources with links

Description Links Data Type Mapping
Height data (DEM - Digital Elevation Model)

SRTM Data Search

SRTM (FTP Download)

Pixel (GeoTiff, Raw) WGS84
Photographical, Vegetation indices (Landsat images)

Global Land Cover Facility

The Gateway to Astronaut Photography of Earth

Landsat image database

Pixel (GeoTiff, MrSID) UTM
Coastline, country borders Digital Chart of the World Vector (SHP) diverse
Whole earth images Earth Images Pixel WGS84
Depth data (bathymetry) DMS: Create Maps and Grids Pixel  
Other interesting links with documentation Download Planet Earth - -

The formats

The standard formats like TIFF, JPG, GIF, etc you might already know. However, there are some wicked compressed formats like MrSID that deal well with large tiled image data. Further, some geographical data is often attached to the images, either in a separate text file or in separate channels, according to extension standards. For example, the GeoTIFF extension stores Latitude, Longitude and mapping in a separate TIFF chunk within the image file. This is very useful when importing the data into GIS (geo information system) software, as you're no longer left with calibrating your corner points. For manual tracing and viewing, MrSID is a nice format, as the appropriate viewers often support fast zooming in. However, for automatical tracing or if you want to obtain specific vegetation indices, you may rather want to use the raw data from the Landsat site. Note that the amount of data to be downloaded is significantly higher as when using a MrSID file.

The mapping problem

When making a map, you may want to have an overlay of all the above data that is interesting to you, like height and vegetation index. Normally, the data comes in different formats, one is mapped in WGS84 (degrees in longitude and latitude), one in UTM (universal transverse mercator), one is orthonormalized on a local coordinate system. You know the problem: How to map a sphere surface to a flat square plane without the least errors. We will not get into the details of projection, but tell you how the problem is basically tackled from a non expert map creator's point of view. A bunch of smart people have solved the problem for you with libraries that are used quite often within several GIS software packages. These libraries are called proj4 and GeoTrans. No worries, you will not have to write your own programs. The next step is a presentation of a selection of software that does the job pretty well.

The software

Program list

These are the programs used to create or extract map data from the various sources:

Table 2. Program list with download links

Program name Deals with format Link
SAGA GIS HGT, GeoTiff, SHP, ... http://www.saga-gis.uni-goettingen.de/html/index.php
GPS Trackmaker SHP, E00, GIF, various GPS data http://www.gpstm.com/
Tatuk GIS Viewer MrSID, GeoTiff, JPG with corner data http://www.tatukgis.com/
GIMP Almost all standard pixel formats http://www.gimp.org/
autotrace GIF, PNG, etc., EMF output http://autotrace.sourceforge.net/

The standard procedure

To first deal with the above mentioned mapping problem, we have to find a common coordinate system to translate the data into. In this case, we use the UTM coordinate systems.

Now this is the overview over the steps you have to take towards a proper map:

  1. Locate the UTM zone ID covering the area you're interested in from the UTM Zone map

  2. Load and convert the DEM (SRTM) data using SAGA GIS - this will be described below.

  3. Generate height curves from the DEM

  4. Overlay this data with the UTM landsat images

  5. Trace the coast lines using GPS Trackmaker (manual tracing) or postprocess with GIMP and autotrace (automatical tracing)

Processing the height data

After you have downloaded the DEM data from the SRTM data sources listed in Table 1, “Data sources with links”, import them into SAGA via Modules>File>Grid>Import>Import USGS SRTM Grid. Note that you may have to load the module DLLs first via Modules>Load Module Library.

Then, you have to do the UTM projection conversion from WGS84 to the UTM zone. Normally, the zone number is automatically derived from the WGS coordinates. Choose Modules>Projection>Grid>Proj4 (Grid) and make sure you have set the parameters in the following dialog according to the values in Table 3, “proj4 to UTM parameters”.

Table 3. proj4 to UTM parameters

Parameter name Value
Target User defined
Direction Geodetic to Projection
Projection Type Universal Transverse Mercator (UTM)
Predefined Standard Ellipsoids WGS84
Unit Meter (1.)

In the following query, the UTM zone will be displayed. If your zone lies on the southern hemisphere, activate the 'South' check box. The last dialog in the conversion process will ask you about the resolution of the resulting grid. The unit is the one specified in the first dialog (default: meters). A grid size of 30 m (per pixel) is a recommended value. You can alternatively choose the resolution of an existing grid by using another Target option.

Last but not least you can generate the contours via Modules>Shapes>Grid>Contour Lines from Grid. The shading is done via Terrain Analysis>Lighting>Analytical Hillshading. As an example result, see Figure 1, “Processing the height data in SAGA”.

Figure 1. Processing the height data in SAGA

Processing the height data in SAGA

Overlay and match

To check whether your converted data matches with a downloaded landsat MrSID image, go through the following steps:

  1. Open the MrSID file in the Tatuk GIS Viewer via Layer>Add.

  2. Zoom into the area of interest and choose File>Export to Image. Use the TIFF format - note that the images can get quite large.

  3. Load the stored TIFF image from within SAGA GIS via Modules>File>Grid>Import>Import Image (...)

  4. When double clicking on the just imported data in the Data tab, you will be asked in what map you want the display in. Use map with the existing UTM grid that was converted in the last step. Now both the datasets will be overlayed. With the Transparency option (displayed in the Parameters tab when selecting the data set in the Maps tab, you can do a proper alpha overlay to check for best match.

Tracing data

There are two ways of tracing the data, either manually or automatically. Manual tracing can be well done from within GPS trackmaker, or you can use existing GPS data from road trips or forest walks.

We will here cover automatical tracing of coast lines and vegetation attributes. To do a proper analysis of that data, we recommend reading of the Landsat Tutorial.

Here is again a quick list of steps you have to take in order to obtain vector data from landsat images. For all image manipulation, we use GIMP.

  1. Combine the interesting Landsat Band channel data into a RGB image according to the Landsat Tutorial using the GIMP option Filter>Colors>Compose.

  2. The resulting RGB image must be separated into a small number of colours for the autotrace process. When converting an RGB image into an indexed image, the number of colours can be reduced. Alternatively, the Posterize option can be used, however, it doesn't often provide optimum results.

  3. Tweak the image, until is shows a good separation of colours. You may also want to use the Filters>Enhance>Despeckle option to get rid of colour noise. Also, manual postprocessing of the image might be necessary.

  4. Run the image through autotrace to obtain an EMF (Enhanced Windows Metafile) vector file. Using autotrace is not very easy, read the autotrace documentation for the meaning of the curve parameter options.

  5. Import the EMF file into GPS Trackmaker for testing.

Combining and publishing the data

Finally we come to the point where you would want to visualize your data. In this example, we use the GPS Trackmaker for output, because it is by far the neatest program for quick editing of data.

The contour lines can be exported from SAGA GIS and be reimported to GPS Trackmaker using the SHP format. For the background image, the already mentioned EMF format is very useful, since the vector data can be rescaled with no pixelization effects and normally takes less memory space. However note that the EMF format does not contain world coordinate information. You will have to calibrate the corner points manually in GPS trackmaker.

Figure 2. Example map in GPS Trackmaker

Example map in GPS Trackmaker

The map can then be printed from within GPS trackmaker, however the printing functionality is somewhat limited. If you want a professional output of the map, you will have to look at tools from ESRI, etc.

Further tips & hints

  • Stay with the GeoTiff format as far as possible, to preserve mapping information

  • Store all data within a SAGA GIS project. Save the parameters often, it is easy to forget what you did.

  • Make sure that you do not bring your high resolution maps into countries where military restrictions apply. This may sound insane, but for example Turkey considers local maps around the resolution of 1:25000 as classified and therefore prohibits production and possession. Ask an insider, before you get yourself in trouble in near-east asian or african countries.

mailaddr


Version: 11.2011