Machine learning-enabled acoustic sensing for RPW infestation detection

Red Palm Weevil(RPW) infestation is a major challenge in palm production, often remaining undetected until severe internal damage has occurred. This proposed work presents a novel non-invasive auditory detection system that combines advanced signal processing with deep learning for the early identification of RPW. Acoustic signals from palm trees were pre-processed to suppress noise, and Linear Predictive Coding (LPC) was used to extract spectral features specific to RPW activity. To refine feature matching, cosine similarity was applied, followed by classification using a Bidirectional Long Short-Term Memory (Bi-LSTM) network. The proposed approach achieved an accuracy of 98.02%, outperforming traditional detection techniques. The novelty of this work lies in integrating LPC based features, cosine similarity, and Bi-LSTM for temporal pattern recognition, enabling highly reliable early detection. A Bi-LSTM RPW classifier combining LPC and cosine similarity, improving accuracy over conventional shallow models was developed. The robust RPW detection system provides accurate outputs despite noise, interference and signal distortion. A limitation of this work is that the evaluation was performed on a controlled dataset, and large scale field validation is required for wider applicability.