Citrus Plant Disease Identification using Deep Learning with Multiple Transfer Learning Approaches

Authors

  • Talha Anwar National University of Computer and Emerging Sciences
  • Hassan Anwar

DOI:

https://doi.org/10.51846/vol3iss2pp34-38

Keywords:

Citrus plant, disease identification, deep learning, transfer learning, classification

Abstract

Citrus plant fruits constitute a significant part of Pakistan's agricultural fruits production. A substantial proportion of citrus fruits is destroyed every year because of different diseases. Citrus plants need to be examined manually to identify the disease prevalence. This paper proposed an in-depth learning approach to identify disease in citrus plants automatically. The dataset used comprised of a small number of citrus plant leaves divided into five categories; 4 of them are disease affected, and the fifth category includes healthy leaves. DenseNet 121 is used as a deep learning model to train the dataset. First, the model is prepared without a transfer-learning approach. Then the model is pre-trained on external data of plant leaves, ImageNet dataset, and a combination of external data and ImageNet data. Model without transfer learning failed to identify the diseases. Model pre-trained on external data resulted in an accuracy of 92%, AUC score of 98.8%, and F1-score of 95%. The model with combined pretraining resulted in an accuracy of 88% and F1-score of 88%. Pre-training on ImageNet data resulted in an accuracy of 82% and F1-score of 87%.

Downloads

Published

2022-04-21

How to Cite

[1]
T. Anwar and H. Anwar, “Citrus Plant Disease Identification using Deep Learning with Multiple Transfer Learning Approaches”, PakJET, vol. 3, no. 2, pp. 34–38, Apr. 2022.

Issue

Vol. 3 No. 2 (2020): Pakistan Journal of Engineering and Technology (Supplementary Issue)

Section

Research Articles

License

COPYRIGHT POLICY

UOL journals follow an open-access publishing policy and full text of all articles is available free, immediately upon acceptance. Articles are published and distributed under the terms of the CC BY-SA 4.0 International License. Thus, work submitted to UOL Journals implies that it is original, unpublished work of the authors; neither published previously nor accepted/under consideration for publication elsewhere. 

Authors will be responsible for any information written/informed/reported in the submitted manuscript. Although we do not require authors to submit the data collection documents and coded sheets used to do quantitative or qualitative analysis, we may request it at any time during the publication process, including after the article has been published. It is author's responsibility to obtain signed permission from the copyright holder to use and reproduce text, illustrations, tables, etc., published previously in other journals, electronic or print media.

Conflict of interest statements will be published at the end of the article. If no conflict of interest exists, the following sentence will be used: "The authors declare no conflict of interest." Authors are required to disclose any sponsorship or funding received from any institution relating to their research. The editor(s) will determine what disclosures, if any, should be available to the readers.

Authors are not permitted to post the work on any website/blog/forum/board or at any other place, by any means, from the time such work is submitted to UOL journals until the final decision on the paper has been given to them. In case a paper is accepted for publication, the authors may not post the work in its entirety on any website/blog/forum/board or at any other place, by any means, till the paper is published in UOL Journals.

The authors may, however, post the title, authors’ names and their affiliations and abstract, with the following statement on the first page of the paper - "The manuscript has been accepted for publication in UOL Journals". After publication of the article, it may be posted anywhere with full journal citation included.

All articles published in UOL journals are open-access articles, published and distributed under the terms of the Creative Commons Attribution-ShareAlike 4.0 International License which permits remixing, transformation, or building upon the material, provided the original work is appropriately cited mentioning the authors and the publisher, as well as the produced work is distributed under the same license as the original.

In the future, UOL may reproduce printed copies of articles in any form. Without prejudice to the terms of the license given below, we retain the right to reproduce author's articles in this way.

Brief Summary Of The License Agreement

By submitting your research article(s) to UOL Journal(s), you agree to Creative Commons Attribution-ShareAlike 4.0 International License which states that:

Anyone is free:

o To copy and redistribute the material in any medium or format
o To remix, transform, or build upon the material for any purpose, even commercially

Provided:

o The author and the publisher have been appropriately credited
o The link to license is provided
o Indicated if any changes were made
o The material produced is distributed under the same license as the original