Discriminating The Corn Plants From The Weeds By Using Artificial Neural Networks

Abstract

The main requisite for weeding-thinning machine is the location of the main stem of the crop. In this study, crops and their positions were
detected using image processing techniques with the aid of artificial neural networks. Morphological operations were performed to singulate
different objects in the images. Several shape features were fed to artificial neural network to discriminate between the weeds and the main crop.
In the final stage, position of the crop was determined which is necessary for the weeding machine to root up all of the other plants. 196 images
consisted of corn plants and four species of common weeds were collected from normal conditions of the field. Results showed that this technique
was able to discriminate corn plants with an accuracy of 100%. It was concluded that high accuracy of this method is due to significant difference
of corns and weeds in the critical period of weeding in the region.