Change Detection and Image Time-Series Analysis 1

9Statistical Difference Models for Change Detection in Multispectral Images

Massimo ZANETTI¹, Francesca BOVOLO¹ and Lorenzo BRUZZONE²

¹Fondazione Bruno Kessler, Trento, Italy

²University of Trento, Italy

9.1. Introduction

For decades, the large number of launched Earth Observation (EO) satellites have provided a unique way to observe our living planet from space. Thanks to the revisiting properties of EO satellites, a large number of multitemporal images are now available in archives. The increasing demand coming from the different application domains has promoted new advances in technology and processing capability to guarantee operational continuity and provide observations for the next generation of operational products, such as land-cover maps, land-use change detection maps and geophysical variables. This allows for an accurate monitoring of the land surface changes in wide geographical areas according to both long-term (e.g. yearly) and short-term (e.g. daily) observations. The detection and understanding of the changes that occur at the ground level is essential for studying global change, environmental evolution and anthropic phenomena. For this reason, the development of change-detection techniques for the analysis of multitemporal remotely sensed images is still one of the most important research topics in remote sensing.

Figure 9.1. A fire occurred in Sardinia Island ...