Multidimensional Digital Filters for Point-Target Detection in Cluttered Infrared Scenes
This work addresses point-target detection in infrared imagery for surveillance or defense applications, presenting an incremental improvement by integrating velocity estimation with existing filtering techniques.
The paper tackled the problem of detecting point-targets in cluttered infrared scenes by using a 3-D spatiotemporal prediction-error filter to enhance foreground/background contrast, achieving improved detection through velocity-based tuning.
A 3-D spatiotemporal prediction-error filter (PEF), is used to enhance foreground/background contrast in (real and simulated) sensor image sequences. Relative velocity is utilized to extract point-targets that would otherwise be indistinguishable on spatial frequency alone. An optical-flow field is generated using local estimates of the 3-D autocorrelation function via the application of the fast Fourier transform (FFT) and inverse FFT. Velocity estimates are then used to tune in a background-whitening PEF that is matched to the motion and texture of the local background. Finite-impulse-response (FIR) filters are designed and implemented in the frequency domain. An analytical expression for the frequency response of velocity-tuned FIR filters, of odd or even dimension, with an arbitrary delay in each dimension, is derived.