Abstract: ‍ ‍The refined characteristics analysis of the rotary-wing drone is the key point to improve the radar's ability for "low， slow and small" target classification. From the perspective of range-azimuth two-dimensional radar imaging， two bands frequency modulated continuous wave radar （FMCW）， i.e.， the K-band and terahertz band， are used to realize the sliding imaging of the rotor drone. Firstly， the signal processing model of radar slide imaging is established， and the realization method of range-azimuth two-dimensional focusing is given. Secondly， the influencing factors of radar imaging performance are analyzed and compared from multiple angles such as radar band， modulation bandwidth， modulation period， slide rail length， imaging time. Finally， three typical rotary wing drones of different sizes were selected to carry out imaging experiments using the two-bands FMCW radars. Analysis of measured data shows that when the drone is at rest， the imaging scattering point can accurately reflect the physical structure of the target； the hovering state is affected by the rotation of the rotor， the echo amplitude is enhanced， and the energy divergence in the distance and azimuth happens， which is related to the blade length and the rotation speed. At this time， micro-motion shows characteristics become more obvious. The properties of the rotor modulation and disturbance on the drone body can be used for subsequent target classification and recognition.