Scholarlink

In Indian culture, medicinal herbs have always been highly valued and seen as important for keeping everyone healthy. Unfortunately, correctly identifying these species is very hard and requires a lot of careful data collection and research. Several types of machine learning methods could change the way medical plants are identified forever. The goal of this paper is to make it clear what role machine learning methods play in quickly and accurately identifying and grouping medicinal plants. Our goal is to give researchers, practitioners, and businesses powerful computer tools for exploring and improving botanical science by combining our knowledge of plants with our knowledge of computers. This project makes the identification process easier while also promoting the protection and study of India's botanical history. The huge variety of Indian medicine plants is shown in our collection, which has 1,835 pictures divided into thirty different groups. We used the latest developments in deep learning tools in the TensorFlow framework. These included Convolutional Neural Networks (CNNs) like MobileNetV2, Densnet121, Xception, and our own custom models. With a test accuracy of 98.64%, MobileNetV2 has done the best of them all.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing Botanical Discovery: Leveraging Machine Learning for Medicinal Plant Identification and Classification

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Sarika Khandelwal,
  • Ruchi Vyas,
  • Harsha Vyawahare,
  • Archana Raut,
  • Jyoti Kumre

摘要

In Indian culture, medicinal herbs have always been highly valued and seen as important for keeping everyone healthy. Unfortunately, correctly identifying these species is very hard and requires a lot of careful data collection and research. Several types of machine learning methods could change the way medical plants are identified forever. The goal of this paper is to make it clear what role machine learning methods play in quickly and accurately identifying and grouping medicinal plants. Our goal is to give researchers, practitioners, and businesses powerful computer tools for exploring and improving botanical science by combining our knowledge of plants with our knowledge of computers. This project makes the identification process easier while also promoting the protection and study of India's botanical history. The huge variety of Indian medicine plants is shown in our collection, which has 1,835 pictures divided into thirty different groups. We used the latest developments in deep learning tools in the TensorFlow framework. These included Convolutional Neural Networks (CNNs) like MobileNetV2, Densnet121, Xception, and our own custom models. With a test accuracy of 98.64%, MobileNetV2 has done the best of them all.