Insect Species Identification Based on Electromagnetic Scattering Parameters Measured by Radar

In this paper, we propose an AdaBoost model for insect species identification based on electromagnetic scattering parameters. Utilizing 14 electromagnetic scattering parameters of insects measured by fully polarimetric radar, an average identification rate of 90.74% is achieved for 9 migratory insect species. Compared with the method that first estimates body size parameters from electromagnetic parameters and then identifies insect species based on those body size parameters, our proposed method directly identifies insect species using electromagnetic parameters and improves the identification rate by 15.55%. This improvement demonstrates the potential for more accurate and efficient automated monitoring systems for migratory pests.