Understanding VisionSketch: Automated Forensic Sketch Generation with Artificial Intelligence

Abstract

The proposed system, VisionSketch: Automated Forensic Sketch Generation Using AI, modernizes the traditional criminal identification process by replacing hand-drawn sketches with AI-powered digital sketch generation. Witnesses or victims can describe or select facial attributes such as eyes, nose, lips, hairstyle, and unique features through a user-friendly graphical interface. Using Python, OpenCV, PIL, and advanced image processing algorithms, the system generates composite sketches in real time. These sketches are stored in a database (MySQL/MongoDB) and compared automatically with existing criminal records, significantly reducing investigation time. The solution minimizes human error, reduces dependence on skilled forensic artists, ensures consistency, and enables scalability. By integrating AI with forensic workflows, VisionSketch enhances speed, accuracy, and reliability, paving the way for future deep learning–based recognition and multimodal biometric integration.