Artifcial Intelligence for Para Rubber Identifcation Combining Five Machine Learning Methods

Abstract

This study aims to identify Para rubber species using a combination of fve machine learning techniquesto classify leaf images. The learning process is defned using a dataset for each classifcation method.Approximately 1,472 leaf images are prepared consisting of various sizes, shapes, quality provided for themodel. The classifcation indicators are defned with the help of an algorithm to identify at least three ofthe top fve potential classifcation outcomes. The algorithm accurately predicts 100% of the fveclassifcation methods. Methods can provide precise and rapid classifcation of large quantities, withoutthe need for image preprocessing prior to classifcation