Research on quality evaluation of maize seed shape based on support vector machine

Abstract

In order to improve the efficiency of maize breeding, the automatic laser cutting robot is applied to maize breeding for sampling of maize seeds. In the process of laser cutting, the quality evaluation of maize seed shape is an important guarantee to realize the high efficiency and automation of sampling. A method for quality evaluation of maize seed shape based on support vector machine is proposed in this paper. Firstly, the images of maize seed are captured and the feature parameters including long axis length, short axis length, area, perimeter and length width ratio of maize seed are extracted by image processing. Secondly, the support vector machine model was established by using these feature parameters as network inputs. Lastly, quality of maize seed shape was evaluated by using linear kernel function, polynomial kernel function and radial basis kernel function. The results show that the quality of maize seed shape can be effectively evaluated by using SVM model of radial basis kernel function. This method is helpful to realize the automation of laser cutting for chip sampling of maize breeding.