Abstract :
This book is a welcome contribution to the growing number of texts in information fusion. The text is primarily focused on the areas in which the author has made direct contributions. This does not mean that the text is narrow in focus; indeed, the author proposes a broad historical, methodological, and algorithmic view of the field. The historical breadth includes a quite enjoyable introductory chapter that traces the origins of information fusion beginning with the seminal contributions of Gauss and Bayes. The discussion quickly brings us to present-day research, and there is a hopeful view of the future and the opportunities that await with the coming Internet of Sensors. There is methodological breadth as the author moves seamlessly between signal processing (level 0 fusion), multitarget multisensor tracking (level 1 fusion), and anomaly detection and situational awareness (higher level fusion). And there is algorithmic breadth. While the author focuses primarily on the multiple-hypothesis tracking (MHT) paradigm, there is a chapter on the expectation maximization (EM) based approach to multitarget tracking: probabilistic multiple hypothesis tracking (PMHT). The author has also conducted research related to the random set approach to tracking, though this paradigm is only mentioned briefly in the text. Chapter 2 takes the reader through the well-known topics of prior information, motion modeling, likelihood functions, and
