Securing Public Places with PCA Based Recognition of Criminal Faces Detected from Surveillance CCTV Footage

Abstract

This paper aims in ensuring the safety of common people at public places by using the existing CCTV systems which are deployed for the surveillance and to determine the security of that place, by identifying the suspicious faces that are captured and notifying the officials. The existing video surveillance systems capture data through CCTVs and store it in their database. After an unpredicted incident already taken place, these databases are used to recognize the culprit. Instead of this, the proposed system keeps a track of the live videos and exacts out frames from it after a fixed interval of time. These frames are then used to fetch faces and compare them with the criminal faces which are already stored in a suspicious faces database, using the feature extraction technique. If the comparison is successful, an alarm is generated which gives an alert about the presence of a criminal at that place. Various face detection algorithms and recognition techniques are used to identify the suspicious face in the crowd and enhances the safety of the public places.