 |
| ||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||
 | |||||||||||||||||||||||||||||||||
 | |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| | |
Today, spatial light modulators (SLM) are used in different microscopic setups. Examples are optical tweezers, programmable phase contrast imaging, confocal imaging, and aberration correction. If the elements are situated in a pupil plane it is possible to use them additionally to correct for aberrations. It is straight forward to correct such aberrations once they are known. Therefore the basic problem is to measure the aberrations. We try to realize this sort of aberration correction without using additional hardware.
Images are typically degraded by aberrations. Most often these aberrations are due to the optical system and it is the optical designer's task to limit them to a suitable - application dependent - amount. This works well for static aberrations but for applications where the aberrations are not constant, dynamic or adaptive correction of aberrations is necessary. The most prominent example is of course the correction of the atmosphere's turbulence for astronomical telescopes.The results that can be achieved are impressive. In other applications, especially microscopy, adaptive optics (AO) also leads to considerable image improvement. Aberration correction for one isoplanatic region and one wavelength is quite easy using an SLM. In this case, just the phase conjugate of the aberrated wavefront, that is the negated wavefront, is written into the modulator. Since typical modulators introduce some unwanted diffraction orders and a pure two pi phase shift often is not easy to achieve one might have to use a carrier frequency to introduce the correction. Anyway, the main problem is to first determine the aberration for applications where this aberration is dynamically changing. The straight forward solution would be to use an addtional wavefront sensor to measure the aberration. Especially with point like objects (laser scanning microscope), this can be easily done. Things are getting much more complicated, if we have an extended object field (wide-field microscopy) and if we do not want to use additional hardware. In the past, we proposed several methods for obtaining such an aberration measurement/correction scheme:
In 1988,
von der Lühe proposed an
interesting method where the correlation of blurred images of different
pupil zones is employed (Optical Engineering 27, 1078-1087, 1988). The
correlation was performed in a
simplified way by optical multiplication of the images with a
mask (corresponding to a pattern within the isoplanatic region).
Different pupil zones are imaged onto different detectors. Today modern
PCs are fast enough to perform the correlations even for
large fields in video-realtime so this optical trick is not
necessary anymore and it is possible to employ digital
correlations. Our method is based on this basic principle and it is
strongly related to the correlating Shack-Hartmann Sensor of Poyneer
(Applied
Optics 42, 5807-5815, 2003) but no additional camera is needed. Fig. 1 shows the basic microscopic setup that we use. It is a video microscope with an additional telescope which images the pupil of the MO onto the spatial light modulator (SLM). Finally the intermediate image plane is imaged onto the CCD camera. In our case we normally use such a setup for programmable phase contrast imaging. For measuring the aberrations we replace the lenslets of
the traditional SHS by gratings with different grating constants and
orientations. Together with the tube lens of the microscope this leads
to multiple (aberrated and bandpass-filtered) copies of the object on
the CCD camera of the microscope. Every copy corresponds to a certain
zone in the pupil. For a perfect plane wave in the pupil the shift of
the copy is directly proportional to the local grating period that we
write into the corresponding pupil zone of the SLM.
 Fig. 1: Principle setup
for measuring and correction aberrations in the microscope by using an
LCD. The pupil of the microscope is imaged onto the spatial
light
modulator by an additional telescope (dotted line)
 Fig. 2: A local tilt of
the wavefront in the pupil (SLM plane) leads to a shift of the
corresponding image For the system we use a Köhler-type illumination with an LED. Different sizes of the local gratings can be used and it is possible to measure 8 different pupil positions simultaneously by spatial multiplexing. The image position due to the local wavefront gradient in the pupil plane is determined currently by correlation. For the wavefront reconstruction we use a linear fit of the obtained gradients to Zernike polynomials. Details about the evaluation, its implementation (Labview), and the optical setup can be found in [4].
We are still in an early stage for using this method and up till now it is not clear how to optimally setup the parameters of the method. Nevertheless, first encouraging results have been obtained for thin specimen. Fig. 3 shows some examples and Fig. 4 shows the comparison of a wavefront before and after detection.  Fig. 3: Some results for
a USAF target  Fig. 4: wavefront before
and after correction Another example:
Fig. 5: Aberration
correction of diatom
© Institut für Technische Optik | ||||||||||||||||||||||||||||||||||||||||||||||||||||
| Letzte
Änderung 09.02.2010
(
 wm) | © Universität
Stuttgart | Impressum |