It is held that the combination of remote sensing and sampling technology can be effectively used in

area measurement. At present, the methods of random sampling, systematic sampling and stratified sampling are

widely used in remote sensing sampling. Based on remote sensing images, this paper has discussed from different

angles the random sampling, systematic sampling, and stratified sampling that includes equal-count, equal-area and

equal-dist. It is found that: (1) for the same features, from three indicators (the average error percentage,

standard deviation and difference), the precision of stratified sampling is better than that of random sampling

and systematic sampling, (2) for different features, the use of three stratified sampling method and the results

obtained are proportional to the percentage of features, i.e., the greater the percentage of features, the better

the results, and (3) from three indicators (the average error percentage, standard deviation and difference), the

three kinds of stratified sampling methods (equal-count, equal-area and equal-dist) have their respective own

merits, and the precision is also closely related to the percentage of features.