International Research Journal of Engineering and Technology (IRJET) Volume: 11 Issue: 10 | Oct 2024
www.irjet.net
e-ISSN: 2395-0056 p-ISSN: 2395-0072
AYURVEDIC PLANT IDENTIFICATION USING DEEP LEARNING Saranya A1, Sowntharya S2, Vaishnavi Deshmukh R3, Kanagadharani K4 1 Assistant Professor, Vivekanandha College of Engineering for Women, Tiruchengode, Tamilnadu (India), 2,3,4
Student Of Vivekanandha College of Engineering for Women, Tiruchengode, Tamilnadu (India). --------------------------------------------------------------------------***----------------------------------------------------------------------training these DL models on extensive datasets containing Abstract—This project aims to leverage the
images of various medicinal plants, our goal is to create a powerful system that accurately classifies plants and provides a wealth of information on their medicinal attributes. [1.2] Key Objectives: • Plant Identification: Implement ResNet50 and VGG19 DL models to achieve precise and efficient classification of medicinal plants based solely on images, ensuring reliability in identification. • Medicinal Content Analysis: Extract and present detailed information on the medicinal contents of each identified plant, offering insights into their therapeutic benefits. • Guidelines for Use: Incorporate age restrictions, genderspecific considerations, and pregnancy restrictions into the DL model, providing users with essential information for the safe and effective use of medicinal plants. • Mode and Dosage Recommendations: Utilize DL to offer insights into the recommended modes of use and appropriate dosage for each detected medicinal plant, ensuring adherence to best practices for optimal health outcomes. [1.3] Impact: This pioneering project aims to revolutionize the identification and utilization of medicinal plants by relying solely on the power of Deep Learning. Through this innovative approach, we seek to empower individuals and healthcare professionals alike, enabling them to harness the healing potential of medicinal plants with precision and confidence, thus contributing to a healthier and more informed society.
capabilities of Deep Learning (DL), specifically ResNet50 and VGG19 models, to address the complex task of classifying medicinal plants. Conventional methods of plant identification often fall short in providing real-time, precise, and comprehensive information about these invaluable botanical resources. The DL-based system developed here aims to accurately classify medicinal plants based on images, offering users detailed information on their medicinal properties. Incorporating factors like age restrictions, gender-specific considerations, and pregnancy restrictions, the model provides essential guidelines for safe and effective plant utilization. Additionally, it offers insights into usage methods and recommended dosages for each identified medicinal plant. By merging advanced DL techniques with botanical expertise, this project aims to create a robust tool for precise plant identification and informed utilization, promoting a more accessible and knowledgeable approach to natural healthcare.
Keywords—Deep Learning Neural Convolutional Neural Network, SVK, Algorithm, Probabilistic Neural Network
Network, Random
I. INDRODUCTION
II. LITERATURE REVIEW
In a world where the healing properties of medicinal plants play a vital role in natural healthcare, the accurate identification and understanding of these botanical treasures are essential. Deep Learning (DL), a subset of Artificial Intelligence (AI), offers a cutting-edge solution to the challenge of classifying medicinal plants. In this project, we leverage the power of DL models, specifically ResNet50 and VGG19, to create a sophisticated system that not only identifies medicinal plants with precision but also provides in depth information on their medicinal contents, age restrictions, gender specific considerations, pregnancy restrictions, mode of use, and recommended dosage. [1.1] The Problem: The vast array of medicinal plants poses a formidable challenge in precise identification and comprehensive information retrieval. Conventional methods often lack the accuracy and efficiency required for real-time and detailed analysis, hindering the widespread and informed use of these invaluable botanical resources for healthcare purposes. Innovative Approach: We propose an innovative approach solely based on Deep Learning, utilizing state-of-the-art models ResNet50 and VGG19. By
© 2024, IRJET
|
Impact Factor value: 8.315
1. Real-Time Identification of Medicinal Plants using Machine Learning Techniques C. Sivaranjani, L. Kalinathan, R. Amutha, R. S. Kathavarayan and K. J. Jegadish Kumar et.al. This research tackles lighting variability in plant leaf segmentation for species identification by employing an enhanced vegetation index, ExG-ExR. Unlike traditional methods relying on fixed thresholds, ExG-ExR's inherent zero threshold adapts to diverse lighting backgrounds, effectively isolating plant regions. Leveraging color and texture features, a Logistic Regression classifier achieves 93.3% accuracy in species classification. This innovative strategy exhibits robust performance, overcoming challenges posed by lighting variations in plant identification.
|
ISO 9001:2008 Certified Journal
|
Page 489