MONITORING THE ABSENCE OF QUEEN BEE IN THE HIVE USING DEEP LEARNING AND HILBERT HUANG TRANSFORM

Abstract

In this paper, we present a fusion method to monitor the absence of the queen bee in a hive using a combination of deep learning neural networks, support vector machine (SVM), and Hilbert Huang transform. First, we collect the sound data from the hive in the presence and missing of the queen bee using the Internet of Things system (IoT). Next, we slice the received audio signal into small chunk with a duration of 10 seconds. In the next step, we perform the Hilbert Huang Transform on each chunks to obtain the spectral image of the audio signal with and without the queen bee. Finally, we use the obtained spectral images to train and test the deep learning neural networks model combined with a support vector machine (SVM) to classify the spectral image of the audio signal with and without the queen bee. The test results on the test set achieved a classification accuracy of 98.61%.