International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 09 Issue: 05 | May 2022
p-ISSN: 2395-0072
www.irjet.net
Ayur-Vriksha A Deep Learning Approach for Classification of Medicinal Plants Soham Kadam1, Premkumar Varma2, Pratiksha Zende3, Prof. Savita Adhav4 1,2,3 B.E
Student, Department of Computer Engineering, G. H. Raisoni College of Engineering and Management, Chas- Ahmednagar, Maharashtra, India - 414005 4 Faculty, Department of Computer Engineering, G. H. Raisoni College of Engineering and Management, ChasAhmednagar, Maharashtra, India - 414005 ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Phytotherapy plays a vital role in maintaining
Historians claim that Ayurveda is a part of Atharva Veda. However, Rig-Veda which is the earliest Veda also mentions about diseases and medicinal plants. This is totally herbal Medicine has no side effect and nearly 10,000 plants used for medicinal purposes. Although herbal medicine does not have any side effects, incorrectly identified medicinal plants may prove fatal to patients. Regardless, identifying a plant's medicinal value is one of the most difficult tasks. At this point, thus, it is necessary to set up an automated system capable of correctly classifying medicinal plants.
the health and well-being of human beings. Identification and classification of medicinal plants are essential to better treatment. However, a lack of expertise in this field severely limits the identification and classification of medicinal plants. This paper proposes Ayur-Vriksha, a Deep Learning Approach, which is based on Convolutional Neural Network (CNN) model and internally it uses Inception V3 and Agile methodology, and for classification of leaf, it uses features like shape, size, color, texture, etc. Ayur-Vriksha helps us to preserve the traditional medicinal knowledge carried by our ancestors and provides an easy way to identify and classify medicinal plants. Our model achieved a classification accuracy of 97% based on our trained dataset, and the proposed dataset contains more than 50 leaf samples of medicinal plants. Finally, the classification is done with Softmax. Ayur-Vriksha allows us to keep the medicinal knowledge passed on from generation to generation.
It may be possible to bridge the gap between a lack of experts and potential requirements in identifying and classifying medicinal plants by using computer vision and image processing methods. Plants are classified by researchers using their shapes, colors, and textures as morphological and spatial features. However, colors are not a useful feature for differentiation because they change throughout the year, as well as differently colored stages within the same leaf. In taxonomy, leaves are classified according to their leaf characteristics. The field has seen a lot of research. It still remains a challenging and unsolved problem due to the high rate of similarity among class members in terms of shape, color, and texture.
Key Words: Classification of Medicinal plant, Collection of Images, Training dataset, Leaf features, Transfer Learning InceptionV3, Agile methodology, Convolutional Neural Network (CNN).
1. INTRODUCTION
2. REVIEW OF LITERATURE
Health is a vital issue for the human race. In recent times, people’s concern regarding health issues has increased exponentially. For developing countries, health care is a fundamental need. Due to the scarcity of doctors and physicians, people of the developing countries have less access to health care services. Thus, health care is a very challenging in these countries. By providing air and water on Earth, plants greatly facilitate the lives and biodiversity of living beings. Medicinal plants, a class of plants that plays a vital role in preventing and treating many diseases, are one of the most important classes of plants. Medicinal plants, one of the important classes of plants, serve as medicines for many diseases on earth by providing air and fresh water. Plants facilitate life by providing air, water, and food. By using pills and tablets (Modern Medicine) the disease will cure very fast but not in depth. India has valuable medicine called ‘Ayurveda’. Ayurveda not only talks about physical health, but also emotional and spiritual health. The origin of Ayurveda dates back to the Vedic era. Most material relating to the health and diseases are available in Atharva Veda.
© 2022, IRJET
|
Impact Factor value: 7.529
To date many researchers have proposed several methods and the fundamental aim of every research is deriving a new solution or invention to solve this problem of identification and classification of medicinal plants. These methods are discussed below in this section. Dileep M. R. and Pournami P.N. [1] proposed a Convolutional Neural Network (CNN) model for classification and the classification is performed using softmax and SVM classifiers. Inception V3 is utilized for the efficient feature extraction from the dataset. This model achieved a classification accuracy of 96.76% and it is tested with leaf samples from 40 medicinal plants. Jing Wei Tan et al. [2] proposed a D-Leaf is a venation-based CNN model, which employs CNN for feature extraction and ANN for classification. Edge detection is used to extract venation details from resized leaf images. The classification accuracy of D-Leaf model is 94.88%.
|
ISO 9001:2008 Certified Journal
|
Page 620