A new micro wedge-ring poor pixels infrared detector is presented, and its special detecting principles are expatiated. The simulated wedge-ring poor pixels images are formed by merging picture elements. Further, with the help of log-polar transformation, it can be easily found that there is much information in the center and little on the edge. Based on grey histogram, targets are detected by using an adaptive real-time threshold segmentation algorithm, and are tracked by adopting a tracking algorithm. In the tracking algorithm the possible states of target were partitioned, and the target position and velocity were estimated on the basis of sequence images. Utilizing the basic geometric figures the wedge-ring poor pixels images are tested. For the different shape basic figures with different orientations, their externalities are mostly different. Even though some have similarities, they can be distinguished by the proportion of pixels’ grey values. The results show that the possible application of poor pixels detector in the target detection and recognition system and it is very significant to continue corresponding research.