Home / Current Issue / Paper 1704475
A Novel Approach to Identify Medicinal Plant Based On Image Processing
Subject area: Science,Engineering and Technology · Area of research: Machine Learning
Abstract
Never forgetting the beautiful things our ancestors did, like giving birth in the home and treating a variety of complex diseases with the aid of nature. They used to be sufficiently knowledgeable about the leaves and their therapeutic benefits, but this generation is falling behind. Our "Dhanvantri" smartphone app, which gives you the medicinal benefits of the leaf by taking a leaf image as an input, was created to aid the community. The two primary functions of Dhanvantri is to recognise a leaf (the input), classify it according to its label, and then retrieve its medicinal characteristics. By comparing the descriptors of each pixel in the image using various layers of convolutional neural networks, we were able to identify the image. The images were then classified based on their labels. At the conclusion, the tagged image's therapeutic properties are revealed. The trained model is ultimately saved in Tensor Flow light format, stored in a database, and invoked using the model reference in Android Studio. To train the model, we used 30 different species of Indian leaves as well as several leaf forms within each species. The outcomes of the Dhanvantri app will aid society in reviving bygone eras.
Keywords
Dhanvantri, therapeutic(medicinal), recognise, classify, descriptors, pixels, convolutional neural network, Tensor Flow lite, leaf detection, bygone eras, arive-dantu, parijata, peepal, relu, roxburgh, basale
How to cite this paper
@article{1704475,
author = {Dandu Sai Satya Vamsi Krishna Raju, Ediga Vinod Kumar, Ediga Mahendra Goud, Dr. Sharmasth Vali Y},
title = {A Novel Approach to Identify Medicinal Plant Based On Image Processing},
journal = {Iconic Research And Engineering Journals},
year = {2023},
volume = {6},
number = {11},
pages = {443-450},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1704475.pdf},
abstract = {Never forgetting the beautiful things our ancestors did, like giving birth in the home and treating a variety of complex diseases with the aid of nature. They used to be sufficiently knowledgeable about the leaves and their therapeutic benefits, but this generation is falling behind. Our "Dhanvantri" smartphone app, which gives you the medicinal benefits of the leaf by taking a leaf image as an input, was created to aid the community. The two primary functions of Dhanvantri is to recognise a leaf (the input), classify it according to its label, and then retrieve its medicinal characteristics. By comparing the descriptors of each pixel in the image using various layers of convolutional neural networks, we were able to identify the image. The images were then classified based on their labels. At the conclusion, the tagged image's therapeutic properties are revealed. The trained model is ultimately saved in Tensor Flow light format, stored in a database, and invoked using the model reference in Android Studio. To train the model, we used 30 different species of Indian leaves as well as several leaf forms within each species. The outcomes of the Dhanvantri app will aid society in reviving bygone eras.},
keywords = {Dhanvantri, therapeutic(medicinal), recognise, classify, descriptors, pixels, convolutional neural network, Tensor Flow lite, leaf detection, bygone eras, arive-dantu, parijata, peepal, relu, roxburgh, basale},
month = {May},
}