HerbId – A medicinal plant identification and recommendation model using machine learning algorithms.
In: AIP Conference Proceedings, Jg. 3122 (2024-05-29), Heft 1, S. 1-9
Konferenz
Zugriff:
Historically, medicinal herbs were the primary means of treating illnesses. However, modern medicine relies heavily on chemical-based drugs, which can be highly effective but often come with unwanted side effects. To address this issue, is to develop a model that can identify medicinal plants for treating various ailments. This model can either take a plant name as input or capture an image and classify the plant as either medicinal or non-medicinal, it can also recommend the plants that can be used to cure the particular symptom entered by the user. If the plant is medicinal, the model provides information about its properties and effects on human body, and its potential cosmetic uses. It has two models, one using supervised and deep learning techniques which is a text-based model called the Plant Name Detection Model [PNDM] and another is an image-based model called the Plant Image Detection Model [PIDM]. For the PNDM model, it is employed using the Logistic Regression algorithm for binary classification. For the PIDM model, gray scaling, Otsu threshold algorithm to automatically learn and extract features from images that are useful for classification. Trained the dataset of plant images, which were labelled as medicinal or non-medicinal, and their associated properties. The output of the model is a prediction of whether the input image or name is medicinal, as well as a list of properties that the model has learned are associated with medicinal plants. The dataset of 300 different plant species is prepared manually which has both medicinal and non-medicinal plants included, for our text-based model. For our image-based model, Flavia dataset from the Kaggle website is used. [ABSTRACT FROM AUTHOR]
Titel: |
HerbId – A medicinal plant identification and recommendation model using machine learning algorithms.
|
---|---|
Autor/in / Beteiligte Person: | Naik, Pavankumar ; Shraddha, K. B. ; Gowda, Sneha Sanjana Dinesh ; Theju, K. V. ; Ruthvika, R. |
Zeitschrift: | AIP Conference Proceedings, Jg. 3122 (2024-05-29), Heft 1, S. 1-9 |
Quelle: | 2024, Vol. 3122 Issue 1, p1-9. 9p.; Jg. 3122 (2024-05-29) 1, S. 1-9 |
Veröffentlichung: | 2024 |
Medientyp: | Konferenz |
ISSN: | 0094-243X (print) |
DOI: | 10.1063/5.0216546 |
Schlagwort: |
|
Sonstiges: |
|