Megha Trivedi1, Abhishek Gupta2* 1 School of Electronics & Communication Engineering, Shri Mata Vaishno Devi University, Katra, Jammu and Kashmir, India
2 School of Computer Science & Engineering, Shri Mata Vaishno Devi University, Katra, Jammu and Kashmir, India
Download Citation:
|
Download PDF
Plants are a source of food, medicines, fiber, fuel, etc. and are therefore crucial for our survival. Due to this, intensive care of plants should be done and it requires monitoring of their growth, size, yield, etc. However, manually monitoring such factors is often time-consuming and necessitates one to have in-depth knowledge of agriculture and plants. Thus, automatic systems for plant image analysis would be beneficial for practical and productive agriculture. Therefore, an automatic method is proposed for monitoring the growth of plants by first performing the segmentation of leaves in plant images and then calculating the segmented area. A deep learning-based architecture “U-Net” was used for the segmentation task. A benchmark dataset of 810 images was used to train and test the proposed deep learning network. The proposed model was trained within 3 hours and achieved a dice accuracy of 94.91% on the training set, 94.93% on the validation set, and 95.05% on the testing set. The proposed architecture was found very lightweight with fewer computations but achieved promising results as compared to other methods in the literature.ABSTRACT
Keywords:
Leaf segmentation, Deep learning, U-Net, Plant monitoring, Computer vision.
Share this article with your colleagues
REFERENCES
ARTICLE INFORMATION
Received:
2020-11-02
Accepted:
2020-12-02
Available Online:
2021-06-01
Trivedi, M., Gupta, A. 2021. Automatic monitoring of the growth of plants using deep learning-based leaf segmentation, International Journal of Applied Science and Engineering, 18, 2020281. https://doi.org/10.6703/IJASE.202106_18(2).003
Cite this article:
Copyright The Author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are cited.