A fuzzy approach to semantic pattern recognition of laser threats is p
roposed. The fuzzy characteristics of feature values of laser threat p
atterns are first described. The computation of membership values for
laser threats are then discussed. Finally, using the error recognition
rate as the criterion, the performances of different aggregation conn
ectives for laser threat pattern recognition are compared. It is shown
that the proposed approach for laser threat recognition can achieve a
n 89% successful recognition rate.