| interferometry_gui - Graphical User Interface (GUI) for SIP functions |
| This is a quick start guide: first of all, this GUI was written to make tests in the photonic field, which means you'll find many functions related to laser images. |
| As it is highly customizable, you can adapt it easily to your needs. |
|
Here is a description of a few experiments and pictures: 1) laser1.jpg is a picture of a laser beam magnified by an microscope objective X10 and filtered by a pinhole of a few microns (=spatial filter). Operation > Profiles show the intensity profiles. LaserBeam > Find Gaussian Profiles allows modelling of these profiles by a gaussian curve (laser beam should have a gaussian profile). The beam waist can be deducted from these values. The laser speckle can be smoothed by one (or several) median filtering(s) found in Operations: this filter removes the high-frequency noise. |
|
2) speckle1.png and speckle2.png are 2 images of an
experiment in speckle interferometry: a Michelson
interferometer is created but the 1st mirror is replaced by
a rugged metallic piece and the second one is replaced by a
metallic rule. A CCD Camera saves a first image. The rule is bent. The Camera take a second picture. Try, Open > speckle1.png then Operations > substract 2 images (absolute value). Fringes appear: between 2 dark fringes, the rule has moved from a distance equal to the light wavelength/2 (here, 633/2=316.5 nm). Normalization, filtering can help better visualization of the fringes. |
|
3) pyramide_wrapped.jpg is what is called a phased image: it
was obtained by projecting fringes on an object then on a
reference plane. The goal is to modelize the object in 3D. Operations > Profiles show that luminance is somewhat proportional to altitude, but each time its value reach 255, there's a jump and the luminance restart from zero. The phase unwrapping process is intended to remove these jumps. PhasedImages > Unwrap linearly can do it with a very good quality image (like pyramide_wrapped.jpg). A more complex algorithm is used by Unwrap by path following: it can unwrap more difficult images. You can test with the merit function called Variance: it's usually the best choice. Be a bit patient when launching those functions: it takes some time (about 2 min on a recent computer). |
| Jocelyn DRUEL <jocelyn.druel1@libertysurf.fr> |
| The latest version of the Scilab Image Processing toolbox can be found at |
| http://siptoolbox.sourceforge.net |
| unwrapp, unwrapl, improfile |