Correction techniques are routinely used to resolve geometric, radiometric, and other problems found in raw remotely sensed data. Another family of image processing techniques is used to make image data easier to interpret. These so-called image enhancement techniques include contrast stretching, edge enhancement, and deriving new data by calculating differences, ratios, or other quantities from reflectance values in two or more bands, among many others. This section considers briefly two common enhancement techniques: contrast stretching and derived data. Later you'll learn how vegetation indices derived from the visible red and near-infrared bands are used to monitor vegetation health at a global scale.
Consider the pair of images shown side by side in Figure 8.16.1. Although both were produced from the same Landsat MSS data, you will notice that the image on the left is considerably dimmer than the one on the right. The difference is a result of contrast stretching. MSS data have a precision of 8 bits; that is, reflectance values are encoded as 256 (28) intensity levels. As is often the case, reflectances in the near-infrared band of the scene partially shown below ranged from only 30 and 80 in the raw image data. This limited range results in an image that lacks contrast and, consequently, appears dim. The image on the right shows the effect of stretching the range of reflectance values in the near-infrared band from 30-80 to 0-255, and then similarly stretching the visible green and visible red bands. As you can see, the contrast-stretched image is brighter and clearer.
Derived data: NDVI
One advantage of multispectral data is the ability to derive new data by calculating differences, ratios, or other quantities from reflectance values in two or more wavelength bands. For instance, detecting stressed vegetation amongst healthy vegetation may be difficult in any one band, particularly if differences in terrain elevation or slope cause some parts of a scene to be illuminated differently than others. The ratio of reflectance values in the visible red band and the near-infrared band compensates for variations in scene illumination, however. Since the ratio of the two reflectance values is considerably lower for stressed vegetation regardless of illumination conditions, detection is easier and more reliable.
Besides simple ratios, remote sensing scientists have derived other mathematical formulae for deriving useful new data from multispectral imagery. One of the most widely used examples is the Normalized Difference Vegetation Index (NDVI). NDVI scores are calculated pixel-by-pixel using the following algorithm:
NDVI = (NIR - R) / (NIR + R)
R stands for the visible red band (MODIS and AVHRR channel 1), while NIR represents the near-infrared band (MODIS and AVHRR channel 2). The chlorophyll in green plants strongly absorbs radiation in the visible red band during photosynthesis. In contrast, leaf structures cause plants to strongly reflect radiation in the near-infrared band. NDVI scores range from -1.0 to 1.0. A pixel associated with low reflectance values in the visible band and high reflectance in the near-infrared band would produce an NDVI score near 1.0, indicating the presence of healthy vegetation. Conversely, the NDVI scores of pixels associated with high reflectance in the visible band and low reflectance in the near-infrared band approach -1.0, indicating clouds, snow, or water. NDVI scores near 0 indicate rock and non-vegetated soil.
Applications of the NDVI range from local to global. At the local scale, the Mondavi Vineyards in Napa Valley California can attest to the utility of NDVI data in monitoring plant health. In 1993, the vineyards suffered an infestation of phylloxera, a species of plant lice that attacks roots and is impervious to pesticides. The pest could only be overcome by removing infested vines and replacing them with more resistant root stock. The vineyard commissioned a consulting firm to acquire visible and near-infrared imagery during consecutive growing seasons using an airborne sensor. Once the data from the two seasons were georegistered, comparison of NDVI scores revealed areas in which vine canopy density had declined. NDVI change detection proved to be such a fruitful approach that the vineyards adopted it for routine use as part of their overall precision farming strategy (Colucci, 1998). Many more recent case studies abound.
The case study described on the following page outlines the image processing steps involved in producing a global NDVI data set.