Digital Criminal Biometric Archives (DICA) and Public Facial Recognition System (FRS) for Nigerian criminal investigation using HAAR cascades classifier technique

In Nigeria, there are many different security concerns and thus crimes have increased despite the fact that there are stringent laws and punishments in place to deter them, making it appear as though the authorities are unable to stop it. In order to identify criminals and conduct investigations, it is imperative that a facial recognition system be connected to a constantly updated digital library. The focus of this paper is to develop an automatic criminal investigation system that can identify criminals based on their faces and produce real-time digital archives about them. However, as an object detection method and facial recognition model, the new system is built on the Haar Cascades Classifier technique in the OpenCV package. Additionally, appropriate programming languages that may provide the needed results were investigated. Python 3.6 was used with the Django 4.2 framework, OpenCV-Python, and Dlib for language execution. Due to Django's ORM, support for numerous databases, and usage of the SQLite3 database, a straightforward database was employed for lightweight applications. The 12 factor app idea was used to construct the DICA-FR system's essential skills. Face detection was applied to the image using the Haar method during processing, and during post-processing, the discovered face was compared with well-known criminal face encodings for matching purposes. Results demonstrated that DICA-FRS could effectively replace human systems since it can recover faces from the furthest distances, display the name of the offender, and sound an alert on the DICA web app's output screen. The DICA system is a working prototype of a system that might be used in the criminal investigative process in Nigeria.