ed using a
ıre (Shortis
ted from a
1ortis et al,
nuch lower
'esulted in
irget image
each of the
each focus
were first
san square)
icrometres,
ilt indicates
zes must be
the quality
a produced
there is a
tion in the
the 16mm
ns exhibits
Infinity
3.0
n)
-1.51
-12.00
-39.92
-92.79
-177.01
he 16mm
1gs.
processed
ork with a
ed in Table
at the near
esent in the
from the
, compared
ower RMS
ce can be
rtion signal
e of target
he network
]uenced by
presence of
latness and
centroid bias, however the clearly evident signal of the
distortion variation is the most likely source of the larger
RMS image error in the near distance network for the
16mm lens.
Lens (mm) 16 20
Focus (m) 1.0 Infinity| 1.8 Infinity
Distance (m) 1.1 3.00 1.8 3.00
Redundancies 6360 8835 | 5490 7785
RMS error (um) | 0.79 0.30 | 0.29 028
Table 2. Results summary for the four networks.
6. FURTHER ANALYSIS
The next stage in the data analysis will be the integration
of the network and straight line data sets. Although an
iterative solution of network and straight line solutions
has been employed successfully, a fully integrated
solution is more efficient (Shortis et al, 1995a). The
integrated approach should improve the accuracy and
independence of the calibration parameters to further
isolate the variation of distortion as a systematic error.
Initial testing of a combined collinearity and straight line
solution within a single network adjustment has
produced encouraging results. The combined far
distance network for the 20mm lens, comprising 4200
target and 3000 straight line observations all at a single
focus distance, realised a slight inflation of the RMS
image error along with à small but significant
improvement in the precisions of the distortion
parameters. The inflation of the RMS image error and
small improvement in the precisions was expected due to
the relatively noisy straight line data. The networks will
be subject to further testing and analysis with re-
measured straight line data.
The final stage in the analysis will be the incorporation
of a number of extended lens models into the combined
network and straight line solution. Distance
interpolation, Brown’s formula and the constant factor
approaches will all be tested using the four networks. In
each case a substantial reduction in systematic error, and
therefore a smaller RMS image error, is expected. The
efficacy of each approach will be compared using
precision analyses of the network solutions and accuracy
analyses against the theodolite coordinate data for the
target array.
7. CONCLUDING REMARKS
This paper has reviewed existing algorithms, described
the experimental design and given preliminary results for
an investigation into the distortion characteristics of the
lenses typically used with still video cameras. Whilst the
539
data reduction and analysis has not progressed to the
stage where definite conclusions can be drawn, it is clear
that significant variation of distortion within the object
space has been detected. It is expected that an extended
lens model will be able to eliminate the systematic error
introduced by the variation of distortion with distance.
The analysis and modelling of this distortion variation
will be reported in a future paper.
8. REFERENCES
Brown, D. C., 1971. Close-range camera calibration,
Photogrammetric Engineering, 37 (8) : 855-866.
Fraser, C. S. and Shortis, M. R., 1992. Variation of
distortion within the photographic field.
Photogrammetric Engineering and Remote Sensing,
58 (6) : 851-855.
Fraser, C. S. and Shortis, M. R,, 1995. Metric
exploitation of still video imagery.
Photogrammetric Record, 15 (85) : 107-122.
Fryer, J. G. and Brown, D. C., 1986. Lens distortion for
close-range photogrammetry, Photogrammetric
Engineering and Remote Sensing, 52 (1) : 51-58.
Fryer, J. G. and Mason, S. O., 1989. Rapid lens
calibration of a video camera. Photogrammetric
Engineering and Remote Sensing, 55 (4) : 437-442.
Magill, A. A., 1955. Variation in distortion with
magnification. Journal of Research of the National
Bureau of Standards, 54 (3) : 153-142.
Shortis, M. R., Burner, A. W., Snow, W. L., and Goad
W. K., 1991. Calibration tests of industrial and
scientific CCD cameras. Invited Paper (Paper 6,
Volume 1), First Australian Photogrammetric
Conference, Sydney, Australia, 12 pages.
Shortis, M. R., Snow, W. L. and Goad, W. K., 1995a.
Comparative geometric tests of industrial and
scientific CCD cameras using plumb line and test
range calibrations. Int. Arch. of Photogrammetry
and Remote Sensing, 30(5W1) : 53-59.
Shortis, M. R., Clarke, T.A. and Robson, S., 1995b.
Practical testing of the precision and accuracy of
target image centring algorithms. Videometrics IV,
SPIE Vol. 2598, pp 65-76
Wiley, A. G. and Wong, K. W., 1995. Geometric
calibration of zoom lenses for computer vision
metrology. Photogrammetric Engineering and
Remote Sensing, 61 (1) : 69-74.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996
