Development of a face recognition system for criminal using image processing and machine learning

The increase in technological developments has reduced the crime rates thanks to its deterrence. The fact that security cameras are very common today has a deterrent feature for people from committing crimes. However, since cameras cannot be used fully integrated with today's technology, this subject has been focused on in this thesis study. Image processing is a method used to transform an image into digital information, perform operations on the image, and some useful information. Face recognition application can be performed using image processing techniques. Through this study, the faces of all criminals with a criminal record will be recorded in a database, and when the criminals who are being searched from these criminals are detected by the security camera at any location, the face of the criminal will be detected through face recognition technology and the camera location will be reported to the relevant police units. Thus, the location of the criminal will be learned by the relevant police unit and it will be possible to be caught by the police. Thanks to this study, it is aimed to prevent criminals with aggressive mind from leaving prison with the intention of revenge from committing crimes again, and to reduce the number of femicides in the world since women are mostly in this situation. However, the most difficult and limiting parts of the study are the problems such as closing the face and seeing half of the face in face detection.