Abstract:
In Rwanda, traditional criminal identification methods, such as thumbprint analysis and witness testimonies, are facing growing challenges as criminals become more adept at evading detection. They increasingly avoid leaving behind fingerprints and commit crimes outside the view of witnesses, limiting the effectiveness of these identification techniques. To address these shortcomings, the use of CCTV cameras in public and private spaces has become more widespread, offering new avenues for surveillance and evidence collection. However, manually reviewing large amounts of CCTV footage is labor-intensive and time-consuming, which limits its effectiveness in real-time criminal identification. This paper proposes an advanced solution in the form of an automated facial recognition system, designed to enhance the identification of criminals using CCTV footage.
The proposed system automatically detects and recognizes faces captured by CCTV cameras, comparing them against a pre-existing database of known individuals. This process leverages the growing deployment of CCTV systems in Rwanda and offers a reliable alternative when fingerprints are absent or witnesses are unavailable. The system works by analyzing facial features and matching them with stored data, thus providing law enforcement with a powerful tool for suspect identification. The automated nature of the system also streamlines the identification process, making it faster and more efficient by reducing the need for manual review of footage.
In terms of methodology, the system is designed to analyze video footage from CCTV cameras to detect faces and extract relevant facial features. Preliminary testing of the system has shown promising results. Around 80% of the input photos have successfully matched with data recorded from CCTV cameras, indicating a high level of accuracy for real-world use. This success rate demonstrates the system’s potential to be a significant improvement over current methods, which rely heavily on physical evidence or eyewitness accounts.
By integrating modern facial recognition technology with existing surveillance infrastructure, this system provides a forward-looking solution to the challenges faced by law enforcement agencies in identifying criminals, particularly when traditional methods are insufficient. Ultimately, the proposed facial recognition system represents a valuable advancement in crime detection and investigation, offering an efficient and reliable tool to enhance the effectiveness of law enforcement operations.
