Title :
Two new categories of shift-invariant features of high-resolution radar range profiles
Author :
Liao, Xuejun ; Bao, Zheng
Author_Institution :
Key Lab. for Radar Signal Process., Xidian Univ., Xi´´an, China
Abstract :
We propose two categories of shift-invariant features for high-resolution radar range profiles. One category consists of the optimally shifted range profiles (OSRP) derived by a method called autoaligning, which aligns range profiles to common reference positions that are flexibly defined on the range axis. The other category, the accumulated bispectral features (ABF), is derived by accumulating the bispectra of the range profiles along circles centered at the origin. The performances of the proposed features are evaluated using a data set of three scaled aircraft models. The experimental results show that the correct aligning ratio of the autoaligning method increases by roughly 100% over the scattering centroid method, and that ABF possess higher separability than FT magnitudes
Keywords :
feature extraction; image resolution; optimisation; radar imaging; radar resolution; radar target recognition; radar theory; spectral analysis; accumulated bispectral features; autoalignment; data set; high-resolution radar range profiles; optimally shifted range profiles; performance evaluation; reference positions; scaled aircraft models; separability; shift-invariant features; Aircraft; Electronic mail; Laboratories; Libraries; Performance evaluation; Radar scattering; Radar signal processing; Shape measurement; Signal processing algorithms; Target recognition;
Conference_Titel :
Signal Processing Proceedings, 1998. ICSP '98. 1998 Fourth International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-4325-5
DOI :
10.1109/ICOSP.1998.770904