C.L. Luengo Hendriks
SPIE vol. 3965, January 2000
Improving resolution to reduce aliasing in an undersampled
C.L. Luengo Hendriks
, L.J. van Vliet
Pattern Recognition Group, Faculty of Applied Sciences, Delft University of Technology,
Lorentzweg 1, 2628 CJ Delft, The Netherlands
We present various methods to increase the spatial resolution of an undersampled, and thus aliased, image sequence. The
sequence is acquired by an infrared camera, which severely undersamples the image of a static scene. Vibration of the setup
causes a random, sub-pixel, global translation of the scene before sampling. Although a single frame is hampered by
aliasing, a sufficiently large series of such frames contains all the information needed to undo this aliasing.
The global translation among the frames is estimated based only on the image content. These shift estimates enable us to
fuse the frames in the sequence, providing a high-resolution image, which is done using interpolation schemes applicable to
non-uniformly sampled data. These interpolation schemes have a built-in regularization to balance resolution enhancement
and increase of signal-to-noise ratio.
The proposed methods are of moderate computational complexity and can be implemented in hardware for real-time use.
They prove to be robust under noisy circumstances, and the performance degrades “gracefully” when the number of input
frames decreases. We compare our results with alternative methods involving iterative algorithms, which are not practical
for real-time application, and a spectral estimation algorithm, which is too sensitive to noise and errors in the shift
Sub-pixel image registration, interpolation, random sampling, aliasing, superresolution.
The lens system in a camera limits the bandwidth of the image projected onto the detector array. The detector array can then
sample this image. If the sampling satisfies the Nyquist criterion, the image projected onto the detector can be exactly
reconstructed from the samples. Detector elements need to be large to collect light efficiently. In infrared imaging, these
elements also need to be isolated from each other, which makes the fill factor small and the pixel pitch large. This causes the
array not to be sufficiently dense to sample according to Nyquist, and the sampled images are aliased.
Some IR camera manufacturers solve this undersampling problem by mounting the detector array on a piezo-electric
element that allows scanning of the image projected onto the detector plane (‘microscanning’). The previous discussion is
also valid for other types of camera; for example, Carl-Zeiss incorporates this technique is the ProgRes digital camera.
The approach explored in this paper is to induce a random motion by vibration of the camera, and combine a series of
frames to create a single image with a higher spatial resolution. Note that this can only work if there is a sub-pixel
translation between those frames. This translation causes the image to be sampled at more points than provided by the