An automatic threshold level selection method for AR marker’s tracking under conditions

Abstract

This paper presents an automatic threshold level selection method for AR marker’s tracking of Augmented
Reality (AR) technology. To compare the efficiency of tracking the AR marker standard, the experiments are
performed under the illumination and distance conditions. Three methods are used including fixed threshold,
global thresholding and automatic threshold level selection. The results under illumination conditions show that the best method is the automatic threshold level selection method which threshold values are direct
variations with illumination of thresholding’s minimum average of 5.92. The results under distance conditions
show that the best method is the automatic threshold level selection method. In this, the distance varies all
ranges of illumination while its distance is inversely proportional to all thresholds. The automatic threshold
level selection method can track all ranges of illumination and obtain the best result compared to other ones.
With the illuminations of 15 lux and 35-45 lux, the most effective distances are 350 mm and 500 mm,
respectively. In addition, the statistical analysis is used to determine the significant differences in the
threshold (p=0.000< α =0.05).