Leaf Vein Fragment Dataset

Published: 20 July 2020| Version 1 | DOI: 10.17632/ngds35s8vr.1
Contributors:
Abdul Hasib Uddin,
,

Description

This dataset contains 64x64, 32x32, 16x16, 8x8, and 4x4 pixel green-channel image sets, each with two classes: 1. Cotyledon-type: Monocotyledon and Dicotyledon. Each cotyledon-type contains 21,828 images. 2. Species: Cocos nucifera, Eichhornia crassipes, Cucurbita moschata, and Neolamarckia cadamba. Each species contains 10,914 images. This dataset contains image data from three types of leaves: 1. Young 2. Mid-aged 3. Old Each leaf has two sides: 1. Front side 2. Reverse side File-format: <cotyledon-type>_<species-first-name>_<leaf-age>_<leaf-size>_<image-height>x<image-width>p.<file-format> Example: ‘Dicot_Cucurbita_midAged_front_32x32p.csv’, which contains 32-by-32 pixel images (flattened to 1024 numerical columns) for the images captured from front-side, mid-aged leaves from Dicotyledon-type of Cucurbita moschata species plant.

Files

Institutions

Khulna University

Categories

Computer Vision, Image Acquisition, Deep Neural Network

Licence