Application of the signal processing technology in the detection of Red Palm Weevil

The Red Palm Weevil (RPW) is the most destructive pest for the date palm in the world. The economic damage to palm crops due to RPW could be mitigated significantly by bioacoustics recognition in an earlier phase of infestation and applying the appropriate treatment. In this paper, a novel signal processing system is developed to detect the existence of RPW through its bioacoustics features. A large set of features are extracted, including some unconventional features which are temporal roll-off, temporal slope, and temporal spread. Since the recorded sound is divided into time frames, an analysis for the criteria behind the choice of the optimum time frame length, as well as the selection of the suitable window function is provided. The results imply the efficiency of the developed system with the selected representative features, window functions, and frame period to detect the existence of the RPW through its feeding sound.