ABSTRACT
Image fusion forms a subgroup of data fusion and uses pixel-based methods. It is very popular in remote sensing for two reasons: (1) Remote sensing acquires diverse, complementary images, and (2) RSIF ensures the combination of the data with the least alteration of their values. This chapter positions pixel-based methods within data fusion, in the neighborhood of feature-and decision-based methods. Originally, three levels were established in the remote sensing community, that is, the pixel, feature, and decision levels. In the meantime, a fourth level, namely, the subpixel level has been introduced. Some researchers include the original signal or in other words sensor as the fourth level as shown in Figure 2.1. It is defined as “the problem of combining multiple measurements from sensors into a single measurement of the sensed object or attribute called the parameter” (McKendall and Mintz 1988). With a quality requirement, the definition is given as: “Sensor fusion is the combining of sensory data or data derived from sensory data such that the resulting information is in some sense better than would be possible when these sources were used individually” (Elmenreich 2002). In remote sensing, signal fusion does not play a major role because the signal is converted into values and images for further consideration. All other levels have their importance in remote sensing. The categorization is necessary to understand remote sensing image fusion as a whole. It is meant to help the reader obtain an idea of the individual characteristics and concerns in each level. In reality, levels are often combined into hybrid fusion because it is most beneficial. In addition, it is sometimes rather difficult to assign a certain approach to one level. It has to be stated that this book was written with a focus on remote sensing image fusion, which refers to the pixel level. The remaining chapters in the book will continue to refer to the levels where appropriate. Last but not least, this chapter contains a section on the value of image and data fusion using remote sensing with other data such as information from a geographic information system, Internet resources, ground truth, and other observations, containing references to higher levels of fusion in remote sensing. The details on the different techniques used in remote sensing image fusion will be described in Chapter 4.
