Gang and moniker identification by graffiti matching

Identifying criminal gangs and monikers is one of the most important tasks for graffiti analysis in low enforcement. In current practice, this is typically performed manually by the law enforcement officers, which is not only time-consuming but also results in limited identification performance. In this paper, we present a system that is able to automatically identify the gang or the moniker for a given graffiti image. The key idea of our system is as follows: given a graffiti query, first find a candidate list of the most similar images from a large graffiti database based on visual and content similarity, and then return the most frequent gang/moniker names associated with the candidate list as the tag for the query graffiti. Our experiments with a large database of graffiti images collected by the Orange County Sheriff's Department in California show that our system is (i) effective in determining the gang/moniker of graffiti, and (ii) scalable to large image databases of graffiti.