Moving a digitizer along a smooth and continuous line could be regarded as a discrete time stochastic process consisting of trend motion and random motion. A stochastic stationary observation series of digitizing error may be generated by adopting a backward difference process (filtering the trend motion from the stochastic series). To separate the trend motion from the stochastic series efficiently, several mathematical formulae have been developed for measuring the complexity of line related to the determination of order of the backward difference operators. The stochastic motion may be simulated by using an autoregressive process in terms of time series analysis theory. The estimation model of digitizing error, consisting of these two processes, has been built. Numerical examples presented in this paper show how to use the model to estimate the digitizing error after having a set of digitized data.
[1]
Gilbert T. Walker,et al.
On Periodicity in Series of Related Terms
,
1931
.
[2]
B. Lavenda,et al.
The statistical-inference approach to generalized thermodynamics
,
1987
.
[3]
Wilhelm Caspary,et al.
Positional accuracy in spatial databases
,
1993
.
[4]
F. Sansò,et al.
Optimization and design of geodetic networks
,
1985
.
[5]
D. B. Preston.
Spectral Analysis and Time Series
,
1983
.
[6]
P. Burrough.
Principles of Geographical Information Systems for Land Resources Assessment
,
1986
.
[7]
George E. P. Box,et al.
Time Series Analysis: Forecasting and Control
,
1977
.
[8]
G. Yule.
On a Method of Investigating Periodicities in Disturbed Series, with Special Reference to Wolfer's Sunspot Numbers
,
1927
.