2026-04-16T21:44:16.133Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:3:11497 2025-07-31T23:31:26Z 3

A Computer Vision-Based Approach Optimized by Modified Metaheuristic for Precise Agriculture Applications, Chapter in AIS Algorithms for Intelligent Systems: ICEAI 2024: Evolutionary Artificial Intelligence, Springer 2025 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/3/11497 https://link.springer.com/chapter/10.1007/978-981-96-5210-5_44 V. Radojcic M. Dobrojevic L. Jovanovic M. Zivkovic M. Mihajlovic N. Bacanin The global demand for food is increasing due to complex factors such as population growth and changing dietary preferences. These factors influence agricultural practices and supply chains. While technological advancements like precision farming contribute to increased productivity, they also raise sustainability concerns. Agriculture, as an intensive sector, is significantly impacted by external factors, particularly weather conditions that can disrupt global supply chains, leading to substantial economic losses. Crop devastation from pests, diseases, and extreme weather events such as droughts and floods further complicates the situation. This paper introduces a modified optimizer based on the particle swarm optimization (PSO) algorithm, specially adapted for optimizing convolutional neural networks for detecting plant diseases. Simulations are conducted on real-world data in order to evaluate the viability of the proposed approach with the best-performing models yielding an accuracy of 89.05%. Key contributions include enhancing accuracy and efficiency through the modified PSO optimizer and developing an AI-based system for precise plant disease detection. bookPart Springer, Singapore 635 647 10.1007/978-981-96-5210-5_44 AIS Algorithms for Intelligent Systems: ICEAI 2024: Proceedings of International Conference on Evolutionary Artificial Intelligence