Abstract: ‍ ‍Accurate estimation of insect morphological parameters is of great significance for insect species identification.At present， the three single-band estimation methods with the highest accuracy include “eigenvalue method”， “the method combined the shape parameter of polarization pattern （i.e.， Radar Cross Section （RCS） of different linear polarization directions） and mean polarization RCS”， and “determinant method”. The first two methods need to judge the parallel/vertical relationship between the maximum direction of the insect polarization pattern and the direction of the body axis （i.e. insect type）. The error of insect type judgment will lead to the increase of estimation error. The “determinant method” does not need to judge the insect type but has poor accuracy. In order to overcome the shortcomings of existing methods and achieve accurate estimation of insect morphological parameters by entomological radar， the characteristics of X-band （9.5 GHz） insect abdominal polarization pattern （i.e.， Radar Cross Section （RCS） of different linear polarization directions） were analyzed. It was found that the minimum value of abdominal polarization pattern （minimum polarization RCS for short） had a good mapping relationship with the body length and body mass of insects， which could be used as a feature to realize the estimation of insect morphological parameters. In order to derive the analytical expression of the minimum polarization RCS， the relationship between the shape of the polarization pattern and the eigenvalue of the scattering matrix was firstly derived based on the symmetric insect model， and then the calculation method of the minimum polarization RCS was derived， and the empirical formula of the estimation of insect body length and body mass based on the minimum polarization RCS was given. The proposed method is named “minimum RCS method”. Compared with the above three methods， it was found that for body length estimation， “the minimum RCS method” had the highest accuracy， while the minimum RCS method has a slightly lower accuracy than “eigenvalue method” and “the method combined the shape parameter of polarization pattern and mean polarization RCS”. The effects of polarization errors on the estimation accuracy of the four methods are analyzed by simulation. The results show that the two methods that need to determine the type of insects are seriously affected by the polarization phase errors， while “the minimum RCS method” has good anti-polarization error performance and is almost unaffected by the phase errors.