Menu Close
  • Home
  • Glossary
  • Recognition
  • Recognition

    Recognition (along with Detection and Identification) is one of three terms commonly used to describe an imaging system task. In this context, recognition means that the observer can differentiate between targets of different classes; for example, between a human target and a vehicle.

    Each of these tasks is best described statistically since these terms historically refer to a human observer’s ability to accomplish the task. How well the tasks can be accomplished in a given situation are described in terms of what fraction of average observers could accomplish it, with the given conditions. For example, “50% Probability of recognition” (sometimes written as P(recog)50%) means that 50% of an average group of observers would say that they can recognize the differences between targets of different classes within the image. According to the Johnson Criteria, it takes 6 pixels across the critical dimension of the target to achieve the 50% level of recognition. See Critical Dimension.