Plant Classification based on Grey Wolf Optimizer based Support Vector Machine (GOS) Algorithm

Abstract

Leaves are the primary identifying feature of trees and other plants. Many of these plants are used in the pharmaceutical industry as industrial crops. Growing automation in industries including commerce and medicine has made accurate leaf identification crucial. Leaves are typically classified according to morphological or genetic characteristics. As a result of their numerous physical differences, however, it is becoming increasingly difficult to categorize the diverse leaf cultivars that exist. Several evolutionary shifts over the past several decades have resulted in an increase in the number of variants of a certain leaf type. To manually sift and identify these leaves is a laborious process. A novel hybrid GOS algorithm is proposed in this study for detecting leaves based on their shape, color, and texture. Three types of leaves (apple, cucumber, and mango) are used as examples, and features for each are extracted using Image Processing techniques, before being optimized with the Grey Wolf Optimizer and finally classified with the SVM (Support Vector Machine) classifier algorithm. Experimental results show that the proposed GOS work improves upon the SVM classifier, with a classification accuracy of 96.83 percent.