PlantML: Some Aspects of Investigation on Deployment of Machine Learning Algorithm for Detection and Classification of Plants

Abstract

Classification and identification of plants are necessary from the perspective of agricultural specialist as well as botanical research. The traditional methods of finding the information for the specific plant consume time and effort. The deployment of machine learning algorithm can play the vital role while identifying as well as classifying the plant. As such, we propose a novel model based on machine learning algorithm that can be deployed to identify the flowers and fruits. We call it PlantML. The proposed work will highlight the experimental arrangement of PlantML as well as the use case, activity diagram of the system. The comparative analysis among applicable machine learning algorithm for PlantML will be discussed. In this work, the deep network knowledge is used to train the datasets considering the features of ImageNet model of deep neural network. The framework platform TensorFlow is utilized to deploy it. The study also highlights that in the domain of image classification, impressive results can be seen while using latest technique of convolutional neural network. The viability of the work will be evaluated to find the evidence that PlantML will be suitable and can act as supplementary tool for agricultural as well as botanical research. As such, from the study, it can be concluded that the proposed model can recognize the different types of flowers and fruits at a higher accuracy.