Vision based guidance of an agricultural combine.

摘要

Agriculture is vitally important to the economy and well being of the United States. Corn and soybeans are two of the most important crops in the Midwest, with 1999 sales of {dollar}17.94 billion and {dollar}12.45 billion respectively (USDA-NASS, 1999). Automated agricultural guidance systems to help produce these crops offering opportunities for reduced operator fatigue, improved safety, increased efficiency and a host of other improvements.; A machine vision based guidance system for agricultural combines was developed at the University of Illinois. Three machine vision guidance algorithms were developed and evaluated for guidance. The first algorithm, the Low Head Mounted Camera Algorithm (LHMCA), utilized images from a head mounted camera that directly viewed the cut/uncut crop edge. The LHMCA was used under actual field conditions in 1999; however, shadows and poor crop condition compromised performance of the algorithm. The High Head Mounted Camera Algorithm (HHMCA) used a camera mounted on the head directly above the cut/uncut edge to image the scene. The HHMCA performed well in the laboratory; poor image quality in the field restricted the use of the algorithm. The Cab Mounted Camera Algorithm (CMCA) processed images from a monochrome image sensor mounted above the cab of the combine. The CMCA closely mimics methodology used by the operator. The CMCA was used to automatically harvest 4.6 ha (12.0 a) of corn during both daylight and at night. The resulting accuracy of the guidance system (13.3 cm) was not statistically different from the accuracy of the GPS position recording system (11.0 cm).