Projects

Bayes

Delaunay

Finger

Homography

Membrane

Multi Cam

Pong

Pool

Robot

Skull

Spykee

Stereo

Wing

Current Projects

ArfArf

OpenCV Qt

OpenCV GTK

OpenCV Python GTK

GTK applet

Spykee

Finger print

(using C) Return
The aim is to locate the finer print endings and forks.

1. Noise deleting

Using a Sequential Switch Filter (FAS Filtre Alterné Séquentiel) and a thresholding.

Original image
Verical components	Horizontal components
Thresholding	Thresholding
Result

As you can see, it is not a good result...

2. Image processing

Original image

An arbitrary threshold (127)

A basic threshold

3. Deformations issues

Using Principal Components Analysis (PCA) (Analyse en Composante Principale ACP).

Deformed image

Result

4. Squelettisation

Using the Hilditch algorithm :

First step : candidates pixels marking for elimination. 4 conditions to be respected 2 < n(p1) < 6 S(p1) = 1 p2*p4*p6 = 0 p4*p6*p8 = 0	Second step : candidates pixels marking for elimination. 4 conditions to be respected 2 < n(p1) < 6 s(p1) = 1 p2*p4*p8 = 0 p2*p6*p8 = 0
p9 p2 p3 p8 p1 p4 p7 p6 p5	p9 p2 p3 p8 p1 p4 p7 p6 p5

Orginal image

Result

5. Caracteristics extraction

What we are looking for :


How we find it :

Fork

Ending

Line

The final result :
Result : thresholing, squeletisation and marked Forks are marked with circles, and endind with squares