Abstract

The relation between an adaptive threshold sensor's size and its performance is analyzed numerically using LMS-filtered noise. The analysis links signal detection to size through the threshold parameter, sampling variance, and a clutter-edge effect. A principle is established for determining by statistical analysis of natural background data whether a 2-D threshold sensor may be advantageous. It is not under the conditions of this simulation. Critical, optimum, and best sizes are given for a 1-D threshold sensor following a 1-D LMS filter with input from an optical sensor having noise equivalent irradiance = 1.5 × 10⁻¹³ W/cm² and spatial resolution of 0.15 × 0.36 mrad.

© 1985 Optical Society of America

Simulation of mid-infrared clutter rejection. 1: One-dimensional LMS spatial filter and adaptive threshold algorithms

M. S. Longmire, A. F. Milton, and E. H. Takken
Appl. Opt. 21(21) 3819-3833 (1982)

LMS and matched digital filters for optical clutter suppression

M. S. Longmire and E. H. Takken
Appl. Opt. 27(6) 1141-1159 (1988)

Statistical detection of resolved targets in background clutter using optical/infrared imagery

Larry B. Stotts and Lawrence E. Hoff
Appl. Opt. 53(22) 5042-5052 (2014)

Noise, clutter, and determinism in infrared spectral signatures under varying meteorological conditions

Dennis F. Flanigan, Alan C. Samuels, Changjiang Zhu, and Francis D'Amico
Appl. Opt. 45(4) 799-808 (2006)

Least-mean-square spatial filter for IR sensors

E. H. Takken, D. Friedman, A. F. Milton, and R. Nitzberg
Appl. Opt. 18(24) 4210-4222 (1979)