Groupe Lab.El - Laboratoire MIPS
Université de Haute-Alsace
IUT Mulhouse
68093 Mulhouse Cedex FRANCE
phone: + 33 (0)3 89 33 76 60
fax: + 33 (0)3 89 33 76 05
o.haeberle@uha.fr
I work on optical properties of imaging systems, more specifically
on the microscope (with all variants). Several image formation models
exist. One can then compute the Point Spread Function of the microscope,
which characterizes its resolution (left picture). Every modelisation
is based on a model and relies on the knowledge of some parameters,
which may differ from their actual values, leading to important differences
between computation and experiments:
Some parameters of the model are easy to measure (thickness and index of the coverslip for exemple), but other are very difficult or even impossible to measure in pratice, as the depth of the specimen under the coverslip, the immersion oil index of refraction (which may vary with temperature), or the effective numerical aperture of the objective. I have shown that it is possible to recover these parameters from a measure of the PSF [1] . This result is important for biologist using microscopes, helping them to precise the experimental protocol of image acquisition. Moreover, one can then use for deconvolution of the image a computed PSF which is noiseless and simultaneously as close as possible to the experimental PSF. One can then use in better conditions the deconvolution algorithms, which are the second topic of the lab. The acquired images suffer from blurring and noise. Deconvolution can (partially) restore the images and correct these defects. We work to improve the algorithms by the automation of the procedure and measurements [2][3]. My work has also lead me to propose a new concept of high resolution
microscopy, called Multiple Objective Microscopy (MOM) [4]. Combining 4Pi and Theta
microscopy, it is possible to achieve a very high resolution (100 nm laterally
and axially) using low numeriacal aperture (0.8) objectives.
If fixed molecules are considered, one must take into account the polarisation moment induced by the excitation PSF, and the reemited dipole field is polarised. The shape and intensity of the PSF then strongly depends on the considered observed polarisation (parallel ort crossed polars) [6]:
|
I worked in this field during my Ph.D. and a post-doc at the Institute for Reference Materials and Measurements (IRMM) of the European Union in Geel-Belgium. I worked (and still work a bit) on physical phenomenons which are candidate for new photonic sources. Specifically , Transition Radiation (Figure (a)), the Smith-Purcell Effect (Figure (b)) and Grating Transition Radiation (Figure (c)). I introduced a new theoretical model, describing Smith-Purcell radiation emitted by an electron moving at an arbitrary angle with respect to the grating rulings (previous models were restricted to trajectories perpendicular to the rulings) [1][2]. Simulations of sources with electrons in the 1 to 100 MeV range have been done, a domain for which no data existed. I have shown the interest of using low energies (E<10 MeV) interacting with millimetric gratings to produce far-infrared radiation. Smith-Purcell radiation is also a candidate to built Free Electron Lasers [3][6]. Experiments have been performed. Two regimes have been considered. When the electrons move quasi-parallel to the grating, the observed radiation has characteristics very similar to Smith-Purcell radiation. Spectra, polarization ans energy dependence have been measured between 20 and 120 MeV [4][7]. When the electrons hit the grating surface at large incidence angles, the emitted radiation has very different characteristics. The radiation has been compared with transition radiation from a flat surface, which has been intensively studied, from both the experimental and theoretical point of views.These experiments have characterized grating transition radiation. A new theoretical model fro transition radiation has been developped, which is also valid for arbitrary surface profiles, contrary to previously existing models [5]. In collaboration with Chiba University, we studied Smith-Purcell radiation from photonic cristals. In particular, we have shown that photonic cristals permit to select very narrow and intense peaks of emission compared to diffraction gratings [6]. 
(a) Transition Radiation (b) Smith-Purcell effect and (c) Grating Transition Radiation Download my PhD (PDF file, zipped) on Smith-Purcell radiation.
|
After my Ph.D., I studied the physical properties of irregular systems at Laboratoire de Physique de la Matière Condensée with Pr. B. Sapoval at the École Polytechnique (France). Fractal geometry gives a rigorous mathematical frame to study irregular objects. This research has direct applications in solid state physics, optics and acoustics. The density of states of irregular systems has been studied,
showing an increase which follows approximately the Weil law. We have also studied the losses in irregular resonators, whith application in acoustics, explaining the decrease of the quality factor which is observed for irregular systems compared to regular ones [2]. The high-frequency regime has been studied in the ray tracing approximation. The distributions of collisions obeys a Lévy law with an exponent a=-2. This could explain some properties of heterogen catalysis in the Knudsen diffusion regime [3]. An experiment to study the vibrational modes of prefractal drums has been carried out [4]. The three above pictures show a computed eigenmode, its simulated hologram and its recorded hologram We have also discovered that fractal arborescences are produced
when burning copper foils (with my collegue B. Keltz): 
|
|
I joined a international team to prove two conjectures relative
to prime numbers:
* I was lucky enough to discover these two Riesel prime numbers, which were the largests at their discovery dates More informations at : www.prothsearch.net
|
I am member of the SETI@HOME program, to analyse data from radiotelescopes. Is there anybody out there?.... I founded the SETI@UHA group.... If you liked Jodie Foster in Contact.... join!
|
