Loading color table (.csv) file to interpret raster data using C#?

Question

I have a raster (GeoTIFF .tif file) that is stored in three bands (Red, Green, Blue), with each band's data type as a byte. I have a separate color table in the form of a .csv file, that stores a height value (in meters), a Red value, Green value, and Blue value on each line. I need to able to utilize this color table to interpret the RGB values in the raster as raw height values. I am using C# bindings of GDAL, OGR and OSR. How do I get the file loaded so that it can be represented as a GDAL color table?

I know OGR has a CSV driver which you can use to Open the file, which I have done. This creates a DataSource object. But I have no idea how to convert this DataSource to a GDAL ColorTable (if it is even possible). Or is another method needed, such as reading the file line by line manually?

Answer 1

Through trial and error I have figured out a working solution to this problem.

1) Load the CSV driver manually:

var csvDriver = OSGeo.OGR.Ogr.GetDriverByName("CSV");

2) Open the CSV file using the driver:

using (var dataSource = csvDriver.Open(fullColorTableFilePath, 0))

3) Get the layer from the DataSource. As far as I can tell this is all of the lines of the file:

var layer = dataSource.GetLayerByIndex(0);

4) Get the first "feature" (line of data) from the file:

var feature = layer.GetNextFeature();

5) Verify the data I need is in the file. I am going to enforce column names with my users in order to ensure consistency and reduce the complexity of the task:

int redIndex = feature.GetFieldIndex("Red");
int greenIndex = feature.GetFieldIndex("Green");
int blueIndex = feature.GetFieldIndex("Blue");
int elevIndex = feature.GetFieldIndex("Start_elev");

6) Get the field values for the line:

var elev = feature.GetFieldAsDouble(elevIndex);
var red = feature.GetFieldAsInteger(redIndex);
var green = feature.GetFieldAsInteger(greenIndex);
var blue = feature.GetFieldAsInteger(blueIndex);

7) Continue reading each line using layer.GetNextFeature.