Classification Of Plants And Weeds In Multi-Leaf Image With Support Vector Machine Based On Leaves Shape And Texture Features

DOI

10.2991/icst-18.2018.171How to use a DOI?

Keywords

Classification of plants and weeds, Leaves shape and texture features, Multi-leaf image

Abstract

We often find plants that have similarities in terms of shape and texture. In agriculture if we want to plant a plant species, other plants will be called weeds as they can inhibit plant growth. We can easily classify plants and weeds using image processing. Watermelon plants were the objects of this research, so other plants besides watermelons will be considered weeds. The recognition of plants based on the similarity of the plant leaves used digital imagery which was divided into three stages. At the first stage, preprocessing was done by cropping the image, resizing the image, separating the background and foreground, and doing edge detection segmentation using the Canny operator. The second stage was feature extraction to retrieve important information for leaf recognition. The features used were features of shape and texture. Then we classified the leaves as leafy plants and weeds using the algorithm of Support Vector Machine (SVM). The SVM method is proven to have good accuracy for classifying plants and weeds based on the shape and texture features in a multi-leaf image using quadratic kernel. The average accuracy is 73.95%.

Copyright

© 2018, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Etriana Meirista
AU  - Imam Mukhlash
AU  - Budi Setiyono
AU  - Dessy Riski Suryani
AU  - Evy Nurvitasari
PY  - 2018/12
DA  - 2018/12
TI  - Classification Of Plants And Weeds In Multi-Leaf Image With Support Vector Machine Based On Leaves Shape And Texture Features
BT  - Proceedings of the International Conference on Science and Technology (ICST 2018)
PB  - Atlantis Press
SP  - 843
EP  - 848
SN  - 2589-4943
UR  - https://doi.org/10.2991/icst-18.2018.171
DO  - 10.2991/icst-18.2018.171
ID  - Meirista2018/12
ER  -