Estimation of Insect Morphological Parameters Based on Single-Frequency Echo Intensity-Independent Estimators

Estimating the morphological parameters of insects, such as mass and length, is crucial for accurate species identification. Traditional entomological radars employ absolute radar cross-section (RCS) estimators to determine these parameters. However, when radars lack angle measurement capabilities, these absolute RCS estimators can underestimate the parameters if insects are not in the beam center. To address this challenge, a method using two single-frequency echo intensity-independent estimators (EIIE) was proposed, but it could only estimate the mass of insects below 200 mg with limited accuracy. This study explores the feasibility of using additional single-frequency EIIEs to estimate the morphological parameters of insects up to 1000 mg. An analysis was performed to examine the relationship between four EIIEs at X-band (9.4 GHz) and the mass and length of 159 insects in an anechoic chamber. The results revealed significant correlations between these EIIEs and the insects' morphological parameters. Subsequently, morphological parameter estimation was achieved by establishing the mappings from four EIIEs to mass and length using random forest algorithm. The findings indicate that this method is effective for insects up to 1000 mg, with mean relative errors of 14.38% for mass and 9.96% for length.