6.1 Introduction

One of the main applications of maritime radar is to the detection of small targets, partially obscured by the unavoidable clutter returns from their environment. In this chapter we attempt to identify procedures that allow this detection to be carried out effectively and consider to what extent we can quantify their performance. The statistical models of clutter and target returns developed in earlier chapters provide us with a framework within which we can address these issues in a reasonably systematic fashion and so define and identify optimal detection procedures. In some simple cases these procedures can be implemented in practice; in others they suggest methods that, while not optimal, are nonetheless well-founded in our understanding of the physical processes involved. This approach can be applied usefully to both incoherent and coherent radar operation. The problem of characterising clutter returns in terms of the K and other statistical models, through the systematic estimation of the parameters that define them, can be addressed in much the same way. The parameter estimation techniques that emerge from this discussion form the basis of useful discrimination algorithms that can be applied to Synthetic Aperture Radar (SAR) and other radar imagery. Many of the concepts introduced in this discussion of detection and discrimination are developed making the assumption that independent samples of the target and clutter returns are available; this simplification is appropriate to the analysis of the detection of small targets in clutter. Nonetheless these same principles can be applied in the discrimination...