A summary of the history of the development of automated remote sensing for agricultural applications

Abstract

Advancing the techniques for machine processing of satellite-acquired multispectral data has been a major thrust of an element of the U.S. civilian remote sensing research program since the 1960's. The program's primary focus has been the use of multispectral data to identify type, condition, and ontogenetic stages of cultural vegetation. The research began as a result of a National Academy of Science study on the applicability of remote monitoring. It was given impetus by the mid-1960s introduction of the first airborne multispectral scanner (MSS) by the University of Michigan. The MSS spawned research by Purdue and others that led to design simulations for the first earth resources satellite in 1969. In 1970, an airborne MSS was used in the Corn Blight Watch — the first large-scale application of remote sensing in agriculture. During 1972 and 1973, research established the feasibility of automating digital classification to process high volumes of Landsat MSS data. The Large Area Crop Inventory Experiment (LACIE) successfully demonstrated automated processing of Landsat MSS data in estimating wheat crop production on a global basis. This experience led to AgRISTARS program designed to address the technical issues defined by LACIE, to investigate other portions of the electromagnetic spectrum and expand the technology to several key commercial crops in important agricultural areas worldwide.