JOURNAL of the JAPANESE SOCIETY of AGRICULTURAL MACHINERY
Online ISSN : 1884-6025
Print ISSN : 0285-2543
ISSN-L : 0285-2543
Distinction between Sugar Beet and Weeds Based on Shape Characteristics Using Image Processing Technique
Masaki TERAWAKITakashi KATAOKAHiroshi OKAMOTOShun-ichi HATA
Author information
JOURNAL FREE ACCESS

2002 Volume 64 Issue 2 Pages 93-101

Details
Abstract

The distinction between sugar beet and three kinds of weeds, green amaranth, wild buckwheat and field horsetail, were tried based on the shape characteristics using the image processing technique. This would be applied to the development of the automatic thinner and weeding machine system for the direct sowing sugar beet. The linear discriminant function, which was based on the six shape characteristics, was produced for the distinction. The correct distinction rate of sugar beet was 85.7%, however, there were distinction errors of sugar beet as green amaranth and wild buckwheat as sugar beet. These error rates resulted in 11.3% and 13.6%, respectively. New characteristics of the angle of the leaf tip portion were introduced to reduce errors. The final results showed that the correct distinction rate of sugar beet was 87.2% and the error was reduced to less than 8%.

Content from these authors
© The Japanese Society of Agricultural Machinery
Previous article Next article
feedback
Top