Adaptive thresholding algorithms maintaining steady false-alarm rates when transitioning between urban buildings, foliage, and rain clutter.
Traditional Cell-Averaging CFAR suffers target masking when multiple closely spaced vehicles enter the reference window. OS-CFAR sorts background cells, selecting a rank-ordered percentile threshold that prevents mutual target suppression.
Stationary urban background clutter (bridges, lamp posts, signs) is modeled in an adaptive spatial background map. Subtracting static clutter profiles isolates moving pedestrians and autonomous vehicles with 25 dB enhanced SNR.