A COMPARISON OF SAMPLING SCHEMAS FOR ASSESSING THE ACCURACY OF CLASSIFIED REMOTELY SENSED DATA

Accuracy assessment determines the value of the resulting information to a particular user, i.e. the information value. In the conventional accuracy assessment an error matrix and some accuracy measures derived from it are used. An error matrix is established using some known samples in reference data and corresponding classified data. There are various factors that affect the performance of the accuracy assessment. The accuracy assessment techniques are not valuable if the method of collecting samples used in error matrix is not considered. One the other factor that affected on accuracy assessment is the number of samples in the map. In this paper with using of four synthetic images having different characteristics and one real image, these factors are verified in some experiments. In these experiments five sampling schema including simple random sampling, cluster sampling, stratified random sampling, systematic sampling, and stratified systematic unaligned sampling are studied. The results represent that depend on specific conditions such as type and size of the study region and object characteristics, different sampling methods and sample sizes are preferred