There are two main ways:
1) Have a third party that is "known correct". This could be someone who has a proven track record, or as is more common in science, a correctly calibrated machine that gives an accurate answer.
2) Have many different surveyors do the job independently with different equipment, and either choose the results that are most common or some kind of statistical average.
Neither of these is a sure fire solution, and in both of them there is always the possibility of error, skew, bias, inaccuracy, etc. Option 1 is either yet another person to wonder about, or a calibrated machine that probably does not exist. Option 2 has the danger that your statistical solution is not really representative (think about the average weight of 3 adults and 3 babies, nobody will be near their average weight).
A variation on option 2 would be to have the same surveyors do the same job several times on different days with different equipment (to avoid equipment bias?). Take the average of each person's results separately and then compare the two people's average results.
Have you googled your question?
Science Buddies Expert Forum