Deep convolutional neural network based plant species recognition through features of leaf
- PDF / 1,428,701 Bytes
- 14 Pages / 439.37 x 666.142 pts Page_size
- 0 Downloads / 204 Views
Deep convolutional neural network based plant species recognition through features of leaf Dhananjay Bisen 1 Received: 29 June 2020 / Revised: 15 September 2020 / Accepted: 6 October 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
In present scenario, the research under image processing has been rapidly transformed from machine learning to deep learning. The deep learning algorithms are usually applied in the various areas like images to be classified or identified more accurately. One of the application areas of deep learning is the plant identification through its leaf which helps to recognize plant species. Botanists consume most of time in identifying plant species by manually scrutinizing and finding its features. This paper proposes an automated plant identification system, for identifying the plants species through their leaf. This task is accomplished using deep convolutional neural network to achieve higher accuracy. Image pre-processing, feature extraction and recognition are three main identification steps which are taken under consideration. Proposed CNN classifier learns the features of plants such as classification of leafs by using hidden layers like convolutional layer, max pooling layer, dropout layers and fully connected layers. The model acquires a knowledge related to features of Swedish leaf dataset in which 15 tree classes are available, that helps to predict the correct category of unknown plant with accuracy of 97% and minimum losses. Result is slightly better than the previous work that analyzes 93.75% of accuracy. Keywords Deep learning . Image pre-processing . Feature extraction . Leaf recognition . CNN classifier . Swedish leaf dataset
1 Introduction There are huge numbers of plants existing in nature and they are one of the life saving element of our planet. Traditionally botanists accord their life to study about plants so that they identify plants species present in earth. But on planet there are about 320 thousand species of plants. It is absolutely impossible task for botanist to study about all these plant species present on
* Dhananjay Bisen [email protected]
1
Rajkiya Engineering College Banda, Atarra, UP, India
Multimedia Tools and Applications
planet and also to find some characteristics which helps them to differentiate the plant species. Bambil et al. [2] collected 40 different kinds of leaves of 30 trees from scanners or phone camera. They identified species using color, shape and texture. Some authors also gave systematic studies on computer vision techniques for plant species identification in terms of plant features. They included approximate 100 peer-reviewed studies, published in last 10 years. Furthermore this study is applied more on available datasets and evaluated classification accuracy [19, 4]. Secondly, Biologist faces a problem in classifying those plant species which are different but they show almost same characteristics. They also invest their huge time on identifying plant species by researching and finding th
Data Loading...
