Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Abstract. Corn is one of the most important grain crops in the world and also one of the three major grain crops in China. However, the level of the mechanization of corn is relatively low in China, especially in the mechanical harvesting. The quality of combine‘s row-follow is the basis of snapping and peeling and threshing. The auto-follow row system can reduce the labor intensity and increase the harvest efficiency compared to the artificial operation. The paper uses the laser radar as the distance sensor to collect data and cluster data by DBSCAN clustering algorithm. Because of the unsatisfied result of traditional DBSCAN clustering algorithm, and the abscissa distance between one point and others in the same row features of corn is proximate, we try to change the Euclidean distance in traditional DBSCAN clustering algorithm to abscissa distance between two points. We took some experiments to test the clustering result. In simulation test, the modified DBSCAN clustering algorithm can detect the boundary correctly and significantly reduce the radius of clustering, and increase the rates of valid data and correctly detect the row features of corn by 13.7% and 15.5%. In field test, the modified DBSCAN clustering algorithm can detect the boundary correctly, the rates of valid data and correctly detect the row features of corn are more than 95%. It provides a theoretical basis for corn combine harvester‘s auto-follow row.