An Efficient Method and Hardware System for Monitoring of Illegal Logging Events in Forest

Illegal logging is one of the major causes of adverse environmental effects such as deforestation, loss of biodiversity, and emission of greenhouse gases. Therefore, illegal logging must be monitored continuously and reported in real time to control such activities. This study introduces a novel wavelet transform based on equivalent-rectangular-bandwidth-scaled gammatone filters for the efficient decoding of audio signals in the time–frequency space, which is further classified using a particle-swarm-optimized support vector machine to identify illegal logging events. The proposed method is prototyped on the hardware system consisting of an audio acquisition unit, a 32-bit embedded platform, a wireless long range (LoRa) module, a LoRa-to-Internet gateway, and cloud and mobile applications to monitor the four classes of illegal logging events and generates a notification in real time. The proposed methodology and system are validated in the laboratory experimental setup that reports overall specificity, accuracy, sensitivity, and F-score; their results are computed as 99.66%, 98.55%, 98.48%, and 98.21%, respectively. Therefore, the proposed system can be suitably deployed for real-time remote monitoring and detection of illegal events in forests.