after the third pruning step resulting in poor measurements
for the direction of the fourth target. The number of
existing targets did not have a major effect on the localization
for the first three dominant targets, with only a minor increase
in the number of outliers.