Automatic detection and recognition of targets by means of passive IR sensors suffer from limitations due to lack of sufficient contrast between the targets and, their backgrounds and among the facets of a target. A set of novel algorithms is designed and tested that uses the target and background Stokes parameters for detection, segmentation, and classification of targets, In these algorithms, it is assumed that for each pixel in the image data three of the four Stokes parameters are provided. This assumption is justified because we have developed a custom designed Polarimetric IR (PIR) imaging sensor that generates three of the Stokes parameters at each pixel location, in real-time. The empirical performance of the above algorithms, in terms of the probabilities of detection, false alarm rate, segmentation accuracy, and recognition probabilities are computed. The comparison of these results with the results associated with intensity-only imaging sensor (nonpolarimetric IR sensory data) shows that use of polarimetric information can noticeably improve the performance.