Multiple-Hypothesis Tracking (MHT)

Multiple Hypothesis Tracking (MHT) is a leading paradigm for Multi-Target Tracking (MTT). In this lecture, we first describe the target and sensor mathematical modeling assumptions that enable track-oriented MHT. Next, we derive the MHT recursion and illustrate its use with illustrative examples. We provide broader context by describing connections between MHT and the overall taxonomy of MTT methodologies. We discuss recent advances in cutting-edge MHT processing and their relevance to operational surveillance applications.