Thursday, September 29, 2016

Visual Biometrics


They're many different methods of gaining biometrics. In the research we're conducting we'll be examining ears and finding ways to detect and extract discriminating features, so the majority of our project will deal with visual biometrics.

To gain some more background on visual biometrics I looked through parts of Matching Shape Sequences in Video with an application to Human Movement Analysis. A paper written by Ashok Veeraraghavan, Amit Roy Chowdhury and Rama Chellappa.

Key things I learned from this paper were:
- Parametric vs Non-Parametric Statistics and models.
- Bioinformatics can be taken from a person's effect on the environment. (i.e. A person's shadow and silhouette can be used as a unique feature for recognition.)

More info on Parametric vs Non-Parametric Stats:
Parametric versus non-parametric

I also started wondering if it was possible to use solely visual biometrics to distinguishing features between twins and came across this paper: A Study of Multibiometric Traits of Identical Twins

Preperation: Getting Setup With OpenCV


For the past week I've been trying to familiarize myself with all the aspects of our research and the resources that we'll need to complete the project.

The primary tool we'll be using for to build our ear classification scheme is OpenCV. Which is a library for performing computer vision related tasks.

Installing OpenCV for Macs isn't as easy straight forward as other platforms like windows.

My research partner and I have followed the tutorials below to get setup with OpenCV:

http://www.mobileway.net/2015/02/14/install-opencv-for-python-on-mac-os-x/

https://jjyap.wordpress.com/2014/05/24/installing-opencv-2-4-9-on-mac-osx-with-python-support/