ACCURACY ESTIMATION OF THE ALGORITHM
OF DIGITAL TRACER IMAGES EVALUATION
based on optimization of the reconstructed image
A.Y.
Mochalin
A new
method for estimation of bias and random errors of computer algorithms of
digital tracer images evaluation is proposed, which, unlike the existing
methods, allows the application of developed in mathematical statistics approaches
for validity analysis of calculated errors. Comparative analysis is carried out
of the influence of
various tracer images parameters on the accuracy of the developed algorithm
based on the reconstructed frame optimization as on the accuracy of the most
popular basic and adaptive cross-correlative algorithms.
Key words: bias
error, random error, probabilistic average, standard deviation, tracer image.