• Md Sefatullah
• Md Shariar Kabir
• Mossaddak Hossain
• Md Parvez Mosaraf

According to FAO (Food And agriculture Organization), there are 11.1% of trees in Bangladesh. And we know that trees benefit us in many ways. After oxygen, food and shelter, the only thing we get is medicine. But there is no proper system to find medicinal plants automatically. This is a big challenge for us. If we can find medicinal plants from 11.1% of the trees in Bangladesh, it will be a great success for us. We have selected a model that is ResNet50 model of deep learning. And we have made a dataset with samples from a total of 6 medicinal plants from a garden in Ashulia, Dhaka, Bangladesh. After that our accuracy came to 99% through this algorithm of deep learning. Also, we have used some algorithm that are VGG16 with 96% accuracy and VGG19 with 97% accuracy. Here we have taken 200 pictures of each leaf of the medicinal plant from different angles. And with the help of these algorithm, we have determined different parameters including height, weight, size, and color of leaves. Then we trained in the deep learning algorithms. After training with some algorithms, our best accuracy is 99% with ResNet50 algorithm. If we can apply this model through our mobile app or web, then everyone will be able to understand by scanning the leaves of the tree through it, which medicinal plant it is.

Plants. Trees. Algorithm. Predictions

"A TRANSFER LEARNING-BASED APPROACH TO DETECT MEDICINAL PLANT IN RURAL AREA", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 6, page no.401-409, June-2022, Available :http://www.jetir.org/papers/JETIRFM06072.pdf

"A TRANSFER LEARNING-BASED APPROACH TO DETECT MEDICINAL PLANT IN RURAL AREA", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 6, page no. pp401-409, June-2022, Available at : http://www.jetir.org/papers/JETIRFM06072.pdf