INTERNATIONAL JOURNAL OF SCIENTIFIC DEVELOPMENT AND RESEARCH International Peer Reviewed & Refereed Journals, Open Access Journal ISSN Approved Journal No: 2455-2631 | Impact factor: 8.15 | ESTD Year: 2016
open access , Peer-reviewed, and Refereed Journals, Impact factor 8.15
Plant Identification in a Combined Imbalanced Leaf Dataset
Authors Name:
P.Pandi selvi
, A.Parameshwari
Unique Id:
IJSDR2304406
Published In:
Volume 8 Issue 4, April-2023
Abstract:
Plant identification has applications in ethnopharmacology and agriculture. Since leaves are one of the distinguishable features of a plant, they are routinely used for identification. Recent developments in deep learning have made it possible to accurately identify the majority of samples in five publicly available leaf datasets. However, each dataset captures the images in a highly controlled environment. This paper evaluates the performance of Efficient Net and several other convolutional neural network (CNN) architectures when applied to a combination of the Leaf Snap, Middle European Woody Plants 2014, Flavia, Swedish, and Folio datasets. To normalize the impact of imbalance resulting from combining the original datasets, the authors used oversampling, under sampling, and transfer learning techniques to construct an end-to-end CNN classifier. Emphasis is placed greater on metrics appropriate for a diverse-imbalanced dataset rather than stressing high performance on any one of the original datasets. A model from Efficient Net’s family of CNN models achieved a highly accurate F-score of 0.9861 on the combined dataset.
"Plant Identification in a Combined Imbalanced Leaf Dataset", International Journal of Science & Engineering Development Research (www.ijsdr.org), ISSN:2455-2631, Vol.8, Issue 4, page no.2636 - 2640, April-2023, Available :http://www.ijsdr.org/papers/IJSDR2304406.pdf
Downloads:
000337212
Publication Details:
Published Paper ID: IJSDR2304406
Registration ID:205593
Published In: Volume 8 Issue 4, April-2023
DOI (Digital Object Identifier):
Page No: 2636 - 2640
Publisher: IJSDR | www.ijsdr.org
ISSN Number: 2455-2631
Facebook Twitter Instagram LinkedIn