Automated Video Surveillance System Using Video Analytics

In this current trend, intellectual video scrutiny is playing a major role in research. It is best suited for various kinds of applications that include security and safety at public areas like airports, traffic control in metropolitan cities and detection of crime that include hit-and-run case and so on. In the hit-and-run case, detection and recognition of object in the video is one of the biggest challenging tasks. Even though several algorithms have been used to detect and recognize the object in the video, it is still considered as a challenging task because of change in illumination. In this work, we have proposed a method for detecting and recognizing a moving object which is involved in crime. We have used Kalman and Gaussian mixture analysis for detecting the object and SIFT and SURF and a classification algorithm like PCA to recognize the object, and continued by the analytics like false acceptance rate and false rejection rate for verification and validating the object during detection as well as recognition process. The investigational resultsshow the efficiency of the algorithms used on various databases. © 2020, Springer Nature Singapore Pte Ltd.