Write Your Own Geospatial Utilities - Open Source Geospatial Tools: Applications in Earth Observation - page 254

14.3.8 The GDAL Algorithms API

In addition to the low level functionalities, such as accessing data, the GDAL API
implements a number of high level functions. For instance, the GDAL Warp API 26
provides high performance image warping functions, a variety of resampling kernels,
and various masking options. The GDAL algorithm API gdal_alg.h implements
a number of algorithms to process raster dataset, including:

• median cut algorithm;

• convert RGB to a paletted image;

• calculate the checksum of an image;

• compute proximity to target pixels;

• fill 'nodata' values;

• polygonize a raster to vectors;

• sieve filter;

• create contour (or iso-) lines;

• rasterize vectors to a raster;

• convert point data to a regular grid.

In most cases you will immediately be able to recognize which algorithm is
implemented by a given function, simply by looking at its name. Many of the GDAL
utilities are using these same algorithms. Because these functions are provided in an
API, it is possible to use them in any of your own utilities as well.

(a) (b)

Fig. 14.11 Landsat 8 image and cloud mask generated with bqa2cloud . a Landsat false color
( R nir, G red, B blue). b Cloud mask

26 http://www.gdal.org/warptut.html