Sensor Evaluation: Resolution Comparison
Satellite sensors are designed with a variety of temporal,
spatial, and spectral characteristics and resolutions. Remotely sensed data may
be acquired on an as needed basis, such as monthly, or hourly, depending on
the frequency of changes of interest. Spectrally, data may be panchromatic
(gray scale imagery or black and white photography), multispectral (several
channels; e.g., blue, green, red, near-infrared), or hyperspectral (hundreds
of channels). The
spectral type
used in a given application depends on the
spectral response
of the observed surface. These two types of resolution are more readily
understood than spatial resolution, which is the physical area of the ground
represented by each picture element, or pixel, in the image. As the pixel
size increases (e.g., 10 to 30 m), the same image size (e.g., 100x100 pixels)
will show a larger area on the ground.
If the pixel size increases and the same area on the ground is displayed,
then the image will get smaller and will have to be magnified to appear
to be the same size.
Examples of how the image area differs with changes in image spatial
resolution are divided into two sets:
1 - 30 m and
30 - 1000 m. The images using
the range of spatial resolutions from 1 - 30 m were derived from a
set of 1-m
ARTIS
(airborne) images. In the range 30 - 1000 m, the images were derived from
a 30-m
Landsat
Thematic Mapper image of the San Francisco Bay area.
Most of the images are multispectral (false color-infrared), with one
panchromatic 1-m image.
Image pixel size changes can be viewed as well as a segment of the
1-m image is
resampled and magnified to 30-m resolution.
Back to the
CHAART Sensor Evaluation page.