Central India Medicinal Plant Dataset (CIMPD).

Abstract

In the present scenario, medicinal plants play a crucial role in promoting a healthy lifestyle by protecting against numerous diseases. They also hold significant potential as a source of income, particularly for rural populations across the globe. Plants used for herbal medicine are known as medicinal plants, and each part of these plants may be utilized for medicinal purposes. Further, medicinal plants are beneficial in enhancing the human immune system. In this research, a new medicinal plant named as Central India Medicinal Plant Dataset (CIMPD) has been developed to support significant research in human health. The dataset contains 9130 leaf images (both healthy and unhealthy) from 23 medicinal plant species. These images were collected from various locations in central India. The entire work was carried out over a period of five months, which included plant selection, leaf collection, image capturing, and data organization into folders. This dataset provides comprehensive information, including the botanical name, common name, geographical origin, healthy and unhealthy leaf images, and medicinal uses of the plants. It serves as a valuable resource for research in machine learning, computer vision, and related domains. Additionally, it will enable the development and evaluation of methodologies for disease detection, plant identification, and other relevant applications.

Plant phenotyping relevance

健康・不健康な葉画像を含む再利用可能なデータセットを構築し、植物の病害状態を画像から判定する研究基盤として提供しているため、画像ベースの植物状態計測に該当する。植物同定も含むが、データセット構築自体が中心である。

Code and data availability

The paper is a data descriptor for the Central India Medicinal Plant Dataset (CIMPD), a public Kaggle dataset of 9130 healthy/unhealthy medicinal plant leaf images from 23 species, directly reproducing the paper's phenotyping (leaf image) measurements. The ResNet18 feature-visualization analysis code is not explicitly,