An improved Model and Estimation Method for Delay/Doppler Altimetry

Semi-analytical model for the echo of spaceborne delay/Doppler altimetry (DDA) has been proved efficient and meaningful. However, the complexity of motion errors could not be neglected in low-altitude and highly maneuver flight, such as unmanned aerial vehicles (UAV) remote sensing. At the beginning of this paper, an improved expression for the flat surface impulse response (FSIR) considering vertical motion errors is proposed. The formulation, being one term of the mean echo power, allows for an explicit understanding of how the various vertical position and slight movement affect the waveform. Then a convolution neural network (CNN) is utilized to classify the objects in quantities which are detected in the first step. Simulated results on a set of delay Doppler maps (DDMs) demonstrate that the proposed method is able to generate desirable description of the scene.