Face Clustering approach for Crime Rate Reduction in Smart City Development

There has been an enormous ascent in the pace of violations in India over the most recent couple of years and reports say that the greater part of these lawlessness are occurred by a gathering of multiple individuals. Subsequently to distinguish and get these lawbreakers, a optimum solution has been proposed. The Emerging field like Convolutional Neural Network is utilized in this arrangement which provides a Face Identification and recognition model. So that the likelihood of relationship of an individual person with other people can be analyzed by this kind of approach. It finds way to monitor a suspected person activity in a social gathering. The framework of this model is the combination of both CNN and AI design. In order to figure out an individual person image, the applications like face identification and face grouping should be done in parallel. This design is suitable for faster compilation and face comparison applications and the results are provided with maximum accuracy compared with eigenfaces approach. The face can be encoded, and it is converted and grouped into a 128-D cluster. And also, the storage of the clusters can be done in the form of array. The confidence relationship can be measured by association rule mining technique that is final step in Grouping algorithm. Application-based interface can be built between the proposed framework and google cloud by which efficient decisions making process is achievable to reduce Crime Occurrence. Because finding the source of Crime is essential in Smart City Development Process.