Title :
Generalization of spatially variant apodization to noninteger Nyquist sampling rates
Author :
Smith, Brian Hendee
Author_Institution :
ERIM Int. Inc., Ann Arbor, MI, USA
fDate :
6/1/2000 12:00:00 AM
Abstract :
Spatially variant apodization (SVA) is reformulated for use on synthetic aperture radar imagery with an arbitrary sampling rate. The algorithm is implemented as a spatially varying three-point finite impulse response filter, and constraints on the filter parameters are developed from physically motivated concepts. By varying the parameters of the filter, the sidelobe energy is reduced with no effective loss of resolution. The procedure produces an output comparable to that of the integer Nyquist version of SVA, and effectively eliminates sidelobe artifacts with no loss of mainlobe resolution
Keywords :
FIR filters; filtering theory; image sampling; parameter estimation; radar imaging; synthetic aperture radar; SAR imagery; SVA; algorithm; filter parameters; finite impulse response filter; mainlobe resolution; noninteger Nyquist sampling rates; sidelobe artifacts elimination; sidelobe energy; spatially variant apodization; spatially varying three-point FIR filter; synthetic aperture radar imagery; Apertures; Convolution; Energy resolution; Extrapolation; Finite impulse response filter; Frequency domain analysis; Image sampling; Radar scattering; Sampling methods; Spatial resolution;
Journal_Title :
Image Processing, IEEE Transactions on