Size Estimation of Spatial Conical Targets Using an Improved Fastest-Descent Method

To optimize radar time resources and reduce computational requirements for target recognition in low-datarate radar echo signals, this paper focuses on the high-resolution one-dimensional range profile sequence of spatial conical targets. A projection model is established, and an improved fastest-descent method is employed for size estimation. The proposed approach first constructs a projection model of the observed target to derive the functional relationship between the projected size and the actual size. Then, the fastest-descent method is used to iteratively solve for the true size of the target. Experimental results demonstrate the effectiveness of the proposed method, achieving a high accuracy in estimating the target's radial length by incorporating precession frequency, with results showing increased stability.