Paper Key : IRJ************977
Author: Akshay Chandrakant Khaple,Abhishek Nitin Deshmukh,Yuvraj Shankar Jha,Pratik Balasaheb Khedkar ,Prof. Aishwarya Yogesh Kadam
Date Published: 01 Apr 2025
Abstract
In forensic science, hand-drawn face sketches are often limited and time-consuming, especially when integrated with modern recognition and identification technologies. This paper presents an innovative Face Sketch Recognition System designed to streamline the identification process in criminal investigations. Our standalone application enables users to create composite sketches of suspects without requiring forensic artists, using an intuitive drag-and-drop interface. These sketches can then be matched automatically with police databases, leveraging deep learning models and cloud infrastructure for efficient processing. The system uses Convolutional Neural Networks (CNNs) for feature extraction and pattern recognition, achieving high accuracy in identifying potential matches. In testing, the model reached an 89% accuracy rate in simulations and an 85% accuracy rate in real-time scenarios, demonstrating its robustness and scalability. Future improvements may involve expanding the facial feature library and enhancing model adaptability for diverse sketch styles, making the Face Sketch Recognition System an accessible and effective tool for law enforcement agencies.
DOI Requested