Stratification and sample selection for multicrop experiments

The author has identified the following significant results. A stratification was performed and sample segments were selected for an initial investigation of multicrop problems in order to support development and evaluation of procedures for using LACIE and other technologies for the classification of corn and soybeans, to identify factors likely to affect classification performance, and to evaluate problems encountered and techniques which are applicable to the crop estimation problem in foreign countries. Two types of samples, low density and high density, supporting these requirements were selected as research data set for an initial evaluation of technical issues. Looking at the geographic location of the strata, the system appears to be logical and the various segments seem to represent different conditions. This result is supportive not only of the variables and the methodology employed in the stratification, but also of the validity of the data sets employed.