We address the problem of detecting vehicle targets on the ground by means of an autonomous polarimetric sensor on board a high-altitude air- or spaceborne platform. We studied a sensor system that incorporates a micropolarizer array, a microscanner, and digital signal processors that host algorithms for detecting small targets. These algorithms use statistical techniques to fuse a target’s Stokes-vector infrared signature measurements. Fusion is achieved by constructing the joint statistical measures for the target’s polarization states. Those states are expressed in terms of the intensity, the percentage of linear polarization, and the angle of the polarization plane. The performance of the sensor system on synthetic polarimetric infrared imagery was evaluated. Relatively good detection and low false-alarm probabilities were achieved even when the number of pixels on targets is as small as four. This is the first report of the development and evaluation of automatic detection in polarimetric imagery containing targets subtending only a small number of pixels. © 2004 Society of Photo-Optical Instrumentation Engineers.