STM Journals Peer-reviewed research from Consortium e-Learning Network

Trends in Opto-electro & Optical Communication

Fingerprint Recognition for Crime Scenes Using Deep Learning

  1. P. Mani Bharadwaj
  2. J. Sai Teja
  3. P. Srivally
  4. B. Sai Reddy

Abstract

Crime-scene fingerprint photos are crucial hints for resolving ongoing cases. Using deep machine learning and convolutional neural networks, we provide a comprehensive crime scene fingerprint identification method in this research (CNN). Precision photography and sophisticated physical and chemical processing techniques are used to collect images from crime scenes, which are then kept as databases. It can be challenging to categorize the photographs taken from the crime scene because they are frequently insufficient. The fingerprint images need to be pre-processed using the appropriate enhancement techniques, and features are then extracted from the fingerprint images. Pre-processed data features are sent into the CNN as training and testing input. According to testing findings on the database using Open CV-Python, it is 80% accurate to identify a partial or complete fingerprint in the criminal database

Keywords

Support