Artificial intelligence approach for diagnosing banana crop diseases through image processing framework: an insightful survey journey from farmers field to digital cloud

Abstract

The significant goal of this article is to highlight the utilization of digital imaging techniques for pathogenic disease detection in banana crops (Musa). The disease occurs in a variety of geographical climatic ranges, posing a threat to banana cultivators worldwide. The inevitable pathogenic invasions in banana plants remarkably influence crop yield loss and consequently agriculture. Subsequently, to enhance cultivation quality and quantity aspects, we involve computer vision models that have been established for automated disease prediction. Moreover, we interpreted the growth evaluation and understanding of different varieties of biotic disease properties that cause damage to the crops. Hence this review aims to address the major challenges in disease detection for identifying disease patterns in stressed plants. It involves classifying infection regions based on different factors such as the type of infectious neighboring plants, and the geometric illusions on the leaf surface area. Additionally, the review addressed the recent strategies of cloud-based computer vision in agriculture which will aid researchers in understanding how digital imaging technologies can be used for pathogen detection in banana plants. Further, this survey ensures a comprehensive review of pathogen diagnosing techniques through artificial intelligence (AI) approaches like convolutional neural networks (CNN), and deep learning (DL) to visually analyze disease patterns based on digital image processing.

Keywords:

• Banana
• plant pathogen
• computer vision
• image processing
• artificial intelligence
• deep learning

Acknowledgment

This paper and the research behind it would not have been possible without the exceptional support of my supervisor, Dr. Sankar Ganesh S, I am also grateful for the insightful support of Dr. Utpal Das, Dr. Rameshkumar S from VIT School of Agricultural Innovations and Advanced Learning for suggesting scientific aspects on plant diseases. We are sincerely thankful to the Vellore Institute of Technology for the core research unit for providing advanced laboratory support.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data sharing not applicable – no new data generated data sharing does not apply as no new data were created or analyzed in this study.