How the Forensic Examiners identify the faces from a crime scene

Hello everyone, welcome back to LetsSeeTech. I hope you all are doing well. Today I am going to share one interesting article How the Forensic Examiners identify the faces from a crime scene. And we will analyze the different regions of human faces in different forensic scenarios. So, guys without further delay lets start.

How the Forensic Examiners identify the faces from a crime scene

When there is a video or image available from a crime scene, for investigation, the most important task for forensic examiners is that face recognition. The Forensic examiners compare images of two faces, they not only focus their attention on the overall similarity of the two faces, but also they focus on the relative distance among different relevant features, Contour of cheek- and chin-lines, Shape of mouth, eyes, nose, ears, etc, Lines, moles, wrinkles, and scars, etc. in the face.

Extraction of Face region:

Forensic examiners carry out facial regions based on three different steps :

  • Detection of facial landmarks.
  • Normalization and facial region extraction.
  • Representation of facial region.

Detection of facial landmarks:

This is the first step to find a facial landmark. This step is divided into two different configurations automatic and manual. Actually in this step first we extract all predefined sets of anthropometric landmarks. Further, for an automatic approach, it automatically detects 65 facial landmarks and for a manual approach, a human manually detects 21 facial landmarks.

See Also: How Technology has affected Humans

How the Forensic Examiners identify the faces

Normalization and facial region extraction:

Once the landmarks have been detected next step is the Normalisation and facial region extraction. So in this step, all the faces were normalized based on ISO standard with an interpupillary pixel distance (IPD) of 75 pixels. This step eliminates rotation and scale in a horizontal plane and eliminates variations in translation and normalized the faces. After the normalization, the next step is the extraction of the facial region.

Representation of facial region:

This is a very popular feature in face recognition. Once the normalization and facial region have been extracted next step is the representation of facial region using Eigen-Region. Here in this step, each face image is divided into different regions, and a noise mask is applied. Then the Eigen region is applied to each facial region.

See Also: Understanding Privacy Perception in Indian Society

Forensic Examiners Databases:

In practice, the forensic examiners carried out all the experimental work using a collection of images of 130 subjects and mugshot from two different databases: One is SCface (which is a static image of human faces) and the second one is a MORPH.

Experimental protocols:

The database was divided into 3 different subsets based on the subject ID: those are development(1-43), SVM training(44-87) and test(88-130). In practice,

  • Mugshot vs mugshot protocol.
  • Mugshot vs CCTV protocol.
  • CCTV vs CCTV protocol.

These protocols are very useful to extract the conclusion so that forensic examiner can improve the face recognition systems. Although there is a lot of research going on focusing on the face recognition system to develop a new technique to bring an improvement to the current face recognition system. The experimental protocols are very useful for forensic examiners because by using these protocols they can improve the face recognition algorithms.

Thank You guys hope this article will help you and if you have any queries please mention in the comment section below we will definitely help you. 


3 thoughts on “How the Forensic Examiners identify the faces from a crime scene”

Leave a Reply

Your email address will not be published. Required fields are marked *