Super resolution techniques have been previously used enhance topographic data directly (instead of via
enhancement of imagery). However, system performance degrades as the pattern database’s size increases. To keep
system performance at an acceptable level, a smaller-than-optimal database must be utilized. This results in error being
introduced due to differences in feature and average region height between the training and presented data. With the
proposed approach, an area-average height value (AAHV) is computed. The work presented compares the averageheight-
and-difference-pattern-based topographic approach to the previous actual-height-based version. The utility of
incorporating an area of pattern overlap is also considered.