Water stress diagnosis in maize crop using enhanced extreme learning machine model for precision irrigation systems

Abstract

Water stress is a significant component that limits crop productivity globally, particularly affecting maize yields. The adverse effects of water stress on maize necessitate an efficient method for rapid and accurate monitoring. An intelligent crop water stress identification model is an important component of the development of a decision support system for smart irrigation.  The present study proposes an intelligent kernel extreme learning machine model (CS-KELM) to recognize water stress in maize crops. To optimize the model's performance, the meta-heuristic cuckoo search algorithm is integrated to fine-tune the model. The proposed approach has demonstrated an accuracy of 94.20% and an F1-score of 94.39%. Integrating the cuckoo search algorithm into the extreme leanring machine (ELM) model has enhanced the model performance, resulting in an improvement of 4.27% in accuracy and a 4.32% increase in F1 score compared to the ELM model. The improved model performance underscores its potential effectiveness in deploying the model into a decision support system for IoT-based irrigation solutions, enabling efficient and precise water delivery based on real-time stress detection