Abstract :

Forensic image analysis plays a critical role in modern criminal investigation by extracting and interpreting physical evidence such as fingerprints, footprints, and bloodstains from images. This work presents a comprehensive framework that integrates advanced image processing and machine learning techniques to detect, enhance, and classify forensic evidence from crime scene photographs. The proposed methodology improves accuracy and efficiency over traditional manual approaches by automating evidence extraction. Experimental results demonstrate robust performance across diverse scenarios and surface types. The research contributes to forensic science by reducing analysis time and improving objectivity. Additionally, the system includes preprocessing, feature extraction, and classification modules designed for real-world forensic application. Real-time evaluation shows significant improvement in evidence detection rates. This paper also identifies future directions for enhancing forens