Objective/significance Based on the urgent needs of the status quo of urban anti-terrorism, combining the present situation of intelligent city construction, we carry out the research on the key problems such as the collection and analysis of counter-terrorism information for urban environment. This paper puts forward a kind of city anti-terrorism information perception system based on the Internet of things, and realizes to improve the level of surveillance and early warning of terrorist activities facing the urban environment. Method/Process The system collects and merges the heterogeneous information of terrorism, which is based on the information perception layer, the intelligence identification layer and the intelligence inference layer, which is based on the city's focal area. To filter suspicious information by constructing a semantic database of terrorist action signals Combined with complex event processing and machine learning technology, real-time monitoring and mining prediction of mass event flow, using the comparison between historical data of terrorist attacks and real-time data, increased the area of counter-terrorism intelligence. Results/Conclusions We hope to provide a better model and an exploratory study on the construction of the counter-terrorism early warning for the urban environment.