Geoscience Reference
In-Depth Information
Chapter 8
Image (Re-)projections and Merging
map projections do convert positions on Earth, typically defined by an angular
(latitude, longitude) pair of coordinates and a geodetic height to a (plane) pro-
jected coordinate system. This chapter deals with raster data operations that are
related to image projections. The main GDAL command line utility for projecting
images is
gdalwarp
. Other utilities related with coordinate reference systems are
gdaltransform
,
gdalsrsinfo
and
gdalmove.py
. Apart from projecting,
gdalwarp
can merge several images into one, which is why we combined both
operations here. Merging is also the subject of
gdal_merge.py
, but this utility
does not support reprojections. For a better understanding of these concepts, we start
by introducing some definitions used throughout this chapter.
8.1 Introduction on Projection and Image Merging
Image
reprojection
transforms an image from one (source) projection to a new
(target) projection. Image
warping
is the process of registering an image with a
georeferenced grid (Fig.
8.1
a). The transformation of image coordinates (row and
column) to georeferenced coordinates (x and y) is based on computing least squares
fit polynomials froma provided set of ground control points (GCP). Creating an image
mosaic
stitches multiple georeferenced input images to a single output image that
covers the bounding boxes of the individual input images (union, see Fig.
8.1
b). If the
images overlap in certain areas, the overlapping pixels values need to be resolved.
This is the subject of image
compositing
(Fig.
8.1
c). Typical rules to compose the
new value are: maximum, minimum, mean and median value. There is currently no
utility in GDAL that deals with image compositing. Current utilities (
gdalwarp
and
gdal_merge.py
) overwrite overlapping pixels with the latest raster. A utility
Search WWH ::
Custom Search