Intelligent crop management system for improving yield in maize production: evidence from India

Sakshi Vishnoi and Jinil Persis

International Journal of Productivity and Performance Management, 2024, vol. 73, issue 10, 3319-3334

Abstract: Purpose - Managing weeds and pests in cropland is one of the major concerns in agriculture that greatly affects the quantity and quality of the produce. While the success of preventing potential weeds and pests is not guaranteed, early detection and diagnosis help manage them effectively to ensure crops’ growth and health Design/methodology/approach - We propose a diagnostic framework for crop management with automatic weed and pest detection and identification in maize crops using residual neural networks. We train two models, one for weed detection with a labeled image dataset of maize and commonly occurring weed plants, and another for leaf disease detection using a labeled image dataset of healthy and infected maize leaves. The global and local explanations of image classification are obtained and presented Findings - Weed and disease detection and identification can be accurately performed using deep-learning neural networks. Weed detection is accurate up to 97%, and disease detection up to 95% is made on average and the results are presented. Further, using this crop management system, we can detect the presence of weeds and pests in the maize crop early, and the annual yield of the maize crop can potentially increase by 90% theoretically with suitable control actions Practical implications - The proposed diagnostic models can be further used on farms to monitor the health of maize crops. Images obtained from drones and robots can be fed to these models, which can then automatically detect and identify weed and disease attacks on maize farms. This offers early diagnosis, which enables necessary treatment and control of crops at the early stages without affecting the yield of the maize crop Social implications - The proposed crop management framework allows treatment and control of weeds and pests only in the affected regions of the farms and hence minimizes the use of harmful pesticides and herbicides and their related health effects on consumers and farmers. Originality/value - This study presents an integrated weed and disease diagnostic framework, which is scarcely reported in the literature

Keywords: Maize; Residual neural network; Local interpretable model-agnostic explanations; Weed control; Pest control; Productivity (search for similar items in EconPapers)
Date: 2024
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.emerald.com/insight/content/doi/10.110 ... d&utm_campaign=repec (text/html)
https://www.emerald.com/insight/content/doi/10.110 ... d&utm_campaign=repec (application/pdf)
Access to full text is restricted to subscribers

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eme:ijppmp:ijppm-11-2023-0620

DOI: 10.1108/IJPPM-11-2023-0620

Access Statistics for this article

International Journal of Productivity and Performance Management is currently edited by Dr Luisa Huatuco and Dr Nicky Shaw

More articles in International Journal of Productivity and Performance Management from Emerald Group Publishing Limited
Bibliographic data for series maintained by Emerald Support ().