Data Processing Tools

Abstract

In this chapter, we discuss the tools and techniques that are conventionally used to process remotely sensed data into geospatial outputs (e.g., classified maps) for further analysis and application. This chapter focuses on the processing of moderate-spatial-resolution, multispectral digital image data, such as the data captured by the Landsat ETM+ sensor. We do not cover the processing of other categories of remotely sensed data such as radar and hyperspectral data; each requires specialized processing approaches that are beyond the scope of this book. Nevertheless, readers interested in learning more about the processing of these other data types should consult additional reading as referenced. Briefly, the processing of multispectral image data is typically divided into three stages: preprocessing, processing, and post-processing. Each of these stages has its own distinct set of common tools and techniques. However, before discussing any of the stages of data processing, it is important to first understand the more fundamental task of image display. Display of Multispectral Image Data Raw image data are most commonly displayed by image processing software in the form of individual bands, or instead, as ''true'' or ''false'' color composites. Any individual band of a multispectral digital image can be displayed as grayscale, where the lowest-value pixels are displayed as black, the highest-value pixels are displayed as white, and pixels with intermediate values are displayed in corresponding shades of gray. Alternatively, true and false color composites will display no more than three image bands at a time, each matched to one of three primary colors: blue, green, and red. As perhaps expected, a true color composite (TCC) image displays the blue band from a raw multispectral image using the blue color ramp, the green band from the image using the green color ramp, and the red band using the red color ramp. Essentially, a TCC depicts its features in natural color. In contrast, a false color composite (FCC) displays the combination of any three bands from a multispectral S. Khorram et al., Remote Sensing, SpringerBriefs in Space Development, DOI: 10.1007/978-1-4614-3103-9_3, Ó Siamak Khorram 2012