With the development of space technology, the number of spacecrafts and debris are increasing year by year. The
demand for detecting and identification of spacecraft is growing strongly, which provides support to the cataloguing,
crash warning and protection of aerospace vehicles.
The majority of existing approaches for three-dimensional reconstruction is scattering centres correlation, which is
based on the radar high resolution range profile (HRRP). This paper proposes a novel method to reconstruct the threedimensional
scattering centre structure of target from a sequence of radar ISAR images, which mainly consists of three
steps. First is the azimuth scaling of consecutive ISAR images based on fractional Fourier transform (FrFT). The later is
the extraction of scattering centres and matching between adjacent ISAR images using grid method. Finally, according to
the coordinate matrix of scattering centres, the three-dimensional scattering centre structure is reconstructed using
improved factorization method.
The three-dimensional structure is featured with stable and intuitive characteristic, which provides a new way to
improve the identification probability and reduce the complexity of the model matching library.
A satellite model is reconstructed using the proposed method from four consecutive ISAR images. The simulation
results prove that the method has gotten a satisfied consistency and accuracy.