Automatic target recognition

Amit Kumar Mishra, Bernard Mulgrew

Research output: Chapter in Book/Report/Conference proceedingEntry for encyclopedia/dictionary


One of the important forms of military intelligence consists of recognizing the type of target in the battle field. If good-quality optical image of the battle scene is available, target recognition is done mostly by human experts. However, for sensors like infrared sensors and radars, target recognition solely by human experts is difficult. Hence, application of pattern recognition techniques, mainly in systems based on sensors like these, is prevalent and is known as automatic target recognition (ATR). This chapter will start with a short description of ATR and the challenges involved. Radar-based ATR, especially synthetic aperture radar (SAR)-based ATR, is one of the most popular ATR schemes, and hence the rest of the chapter will be devoted to SAR ATR. We will first describe the different stages of a generic ATR system and then elaborate some of the major challenges of radar-based ATR. This will be followed by a description about the major figures of merit for a SAR ATR algorithm. The chapter will discuss some of the important algorithms from the open literature dealing with radar-based ATR. We will also discuss some of the recent developments in the field of ATR, for example, bistatic radar-based ATR, polarimetric radar-based ATR, and multi-look radar-based ATR. The chapter will close with a note on the futuristic radar-based ATR systems.
Original languageEnglish
Title of host publicationEncyclopedia of Aerospace Engineering
ISBN (Electronic)9780470686652
ISBN (Print)9780470754405
Publication statusPublished - 15 Dec 2010
Externally publishedYes


  • confusion matrix
  • template matching
  • Bayesian classifier

Cite this