FORENSIC FACE SKETCH CONSTRUCTION AND RECOGNITION

Abstract

The overall crime rate is rising daily in the modern day, and in order to keep up, law enforcement agencies must also discover ways to streamline investigations and aid in bringing offenders to justice. Utilizing face recognition technology to identify and confirm the culprit is one such method. The conventional method in this case is to utilize the hand-drawn face sketches created by forensic sketch artists to identify the offender; modernizing this would entail using the hand-drawn sketch and then comparing them with the database of the law enforcement agency to identify the offender. This method would have a number of technological restrictions and even take a lot of time because there are so few criminalists who draw sketches accessible compared to the increased crime rate.

By giving the law enforcement department a stand-alone platform that would enable users to generate an accurate face sketch of the suspect without a forensic sketch artist's assistance, particular training, or creative abilities, this initiative aims to shorten the duration and speed up the procedure. With a range of face elements, the sketch may be made using the drag-and-drop functionality of the program. The created composite face sketch is then automatically compared with the database of the law enforcement agencies using deep learning.