Abstracto

A research on the airport monitoring intelligent recognition of terrorist violent actions

Zhang Ning, Zhu Jin-fu

It’s very important to identify terrorists’ behavioral features for the sake of airport security. Generally, the traditional method is targeted at personnel recognition of specific features of fixed scenes. The nonrestrictive characteristic of airport regions and the disguise of violent terrorists are left unconsidered. As a result, the recognition evolves into a high dimensional issue with increased difficulty in recognition. This paper presents a projection feature algorithm to identify the behavioral features of terrorists in an intelligent way. The researchers reconstruct the features of the original collected monitoring images. During reconstruction, the optimal cost function is selected to provide a proper data basis for accurate recognition. The reconstructed two-dimensional images are projected on the one-dimensional plane and transformed into the one-dimensional recognition issue. The recurrence relation is applied to categorize and recognize related violent terrorist actions with enhanced recognition accuracy and efficien