Agriculture Reference
In-Depth Information
Landsat captures images of a region every 16 days as a result of only several useful
images within a year, whereas more images can be provided by aircraft remote sens-
ing used for the analysis of rapid changes of seasons. Therefore, both satellite and
aircraft remote sensing are necessary for development of agricultural remote sens-
ing. The Jet Propulsion Laboratory developed a hyperspectral image sensor named
the Airborne Visible Infrared Imaging Spectrometer (AVIRIS), which covers the
range of 0.4-2.5
m in 224 contiguous bands with approximately 10 nm bandwidth
and has a spatial resolution of 20 m. Similar to AVIRIS, the Compact Airborne
Spectrographic Imager is a hyperspectral image sensor, which has features of a push-
broom charge-coupled device (CCD) with high spatial resolution from 0.5 to 10 m
and an adjustable spectral range between 400 and 1000 nm of up to 288 program-
mable spectral bands at 1.9-nm intervals. The Airborne Imaging Spectroradiometer
for Application, developed by Spectral Imaging Ltd., is a hyperspectral image sensor
with a 2-m spatial resolution in the wavelength range of 0.43 to 1.0
μ
m with 512 spec-
tral bands as the maximum. The Airborne Digital Sensor (ADS40), developed by
LH Systems and the German Aerospace Center, has forward-, nadir-, and backward-
looking linear CCD arrays for providing high spatial-resolution panchromatic
images. The multispectral data with a high spatial resolution of 20 cm acquired by
the ADS40 has blue channel (B) in the spectral wavelength range of 0.43 to 0.49
μ
μ
m,
green channel (G) in 0.535 to 0.585
μ
m, red channel (R) in 0.61 to 0.66
μ
m, and NIR
channel in 0.835 to 0.885
μ
m, and can be used for applications in crop and land use
analysis.
Aerial photographs are most useful for the applications when fine spatial detail
is more critical than spectral information, because the spectral resolution of such
images is generally not as good as the data captured with electronic sensing devices
(Ahamed et al., 2011). The well understands of geometry of vertical photographs
makes very accurate measurements available for a variety of applications in geology,
forestry, and mapping. Photogrammetry is defined as the science of making measure-
ments from photographs, and has been widely performed since the very beginning
of aerial photography. Currently, aerial hyperspectral images are also applied for
agricultural remote sensing (Yang et al., 2002; Yao and Tian, 2004). Hyperspectral
imaging can acquire images containing more bands (tens to hundreds) with a nar-
row band (one to several nanometers) in the same spectral range as a multispectral
image (Schowengerdt, 1997). Therefore, more detailed information is available on
the basis of hyperspectral image data (Yao and Tian, 2004). Information on crop
fields is important to agricultural application, and although field maps are available
for most farms, they are commonly not available in a computerized or digital form.
Digital maps can identify crop areas falling within a selected region by using GIS.
There are three ways to obtain accurate digital field maps: GPS (global position-
ing system) surveys, digital orthophotos, or by digitizing field boundaries. A digital
orthophoto is composed of a series of aerial photographs that have been mathemati-
cally stretched and joined to generate a single computerized photograph accurately
in latitude and longitude as well as in elevation (Gers and Erasmus, 2001).
High-resolution aerial photographs have several important rangeland appli-
cations, such as monitoring of vegetation change, development of grazing strate-
gies, determination of rangeland health, and assessment of remediation treatment
