323.

[Cover] [Contents] [Index]

Page 9

(1.1)

This index is based on the observation that vegetation has a high reflectance in the near-infrared band, while reflectance is lower in the red band (Figure 1.5). NDVI is less sensitive to the changes of atmospheric conditions than are other indices (Jackson, 1983; Holben and Kimes, 1986) and therefore has been widely applied for vegetation monitoring. Data from other sensors that provide red and near-IR images can also used to generate NDVI images. For instance, in the case of Landsat TM images, NDVI is based on bands 5 and 7, while for SPOT HRV, NDVI is derived from bands 3 and 2.

Reflectance from water surfaces is relatively low, and is more or less zero at wavelengths beyond the visible red. Knowledge of surface material reflectance characteristics provides us with a principle on the basis of which suitable wavebands to scan the Earth’s surface for a particular mission can be selected (e.g. for vegetation monitoring, sea surface observation, or lithological identification). Such knowledge also provides an important basis to make the objects more distinguishable in terms of multiband images manipulations such as overlay, subtraction, or ratioing.

1.1.3 Spatial and radiometric resolution

The resolution of a remote sensing instrument can be expressed in terms of its spatial and radiometric resolution. The higher the spatial resolution the smaller the ground objects that can be distinguished. The spatial resolution is related to the instantaneous field of view (IFOV) of the sensor, which denotes the size of the area from which the sensor receives the energy at the given instant in time. In Figure 1.6 the energy transmission path to the sensor takes a cone-like shape, and ground resolution is roughly equivalent to the diameter of the circle formed by the intersection of this cone and the ground surface. As the sensor scanning area moves away from the nadir, the larger the IFOV will be (Figure 1.6), and thus results in the distortion of the resulting image as the spatial scale decreases from the left and right edges of the image towards the centre. The correction for this scale distortion effect (and other effects due principally to the platform’s orbit characteristics and to the eastwards rotation of the Earth during scanning) is called geometric correction. This correction is usually performed by constructing a transform (using either an empirical procedure based on least squares methods or an analytical procedure using orbital information) in order to map ground co-ordinates to their corresponding image co-ordinates, and vice versa. A review of local and global methods of geometric correction and image registration is provided by Brown (1992).

One’s instinctive feeling might be that the finer the spatial resolution the

[Cover] [Contents] [Index]