Insect wing‐beat frequency automatic extraction and experimental verification with a Ku‐band insect radar system

Abstract

The wing-beat frequency is one of the most important behavioural parameters of migratory insects and is widely used to distinguish insect species due to the diversity of wing-beat frequency among different species. Traditional insect radar has been able to achieve automatic extraction of the wing-beat frequency. However, its low-range resolution and fixed threshold detection may cause false alarms and reduce the precision of wing-beat frequency. The authors� Ku-band insect radar system enables high-resolution, full-polarised detection of insects, but has not achieved automatic extraction of wing-beat frequency yet. The authors propose an insect wing-beat frequency automatic extraction algorithm based on a Ku-band insect radar system. The algorithm adopts the constant false alarm rate detector and tracks the association algorithm to get target echo automatically and combines the amplitude spectrum method and the phase spectrum method to achieve wing-beat frequency extraction. The authors carried out the outfield experiment using the radar system and used the collected data to test the proposed algorithm. The experiment results confirm that the algorithm can extract the wing-beat frequency of insects automatically based on the Ku-band insect radar system, which can improve system efficiency.