<title>Synthetic vision using polarization-sensitive thermal imaging</title>

Abstract

Landing of aircraft in inclement weather and taxiing operation in the presence of copious obstacles is a major issues in air traffic control for both military and civilian aviation. Onboard sensors are needed to penetrate smoke, fog, and haze and to provide enough resolution for the automated detection and recognition of runways and obstacles. The performance of automatic target recognition (ATR) systems using thermal infrared (FLIR) images is limited by the low contrast in intensity for terrestrial scenes. We are developing a thermal imaging technique where, in each image pixel, a combination of intensity and polarization data is captured simultaneously. Images of polarization have useful contrast for different surface orientations. This contrast should facilitate image segmentation and classification of objects. In this paper, we will describe a combination of two innovative technologies: a polarization-sensitive thermal imaging sensor and a suite of polarimetric specific automatic object detection and recognition algorithms. The sensor has been able to capture polarization data from thermal emissions of automobiles. Surface orientations can be measured in the same image frame as temperature distribution. For the evaluation of the algorithms a set of performance metrics will be defined. We will discuss our evaluation of the algorithms on synthetic images as would be captured with the polarization-sensitive sensor. We will compare the polarimetric specific ATR performance with the performance of conventional FLIR-based ATR.