Comparison of the Maximum Likelihood Ratio Detectors with the Hough Detectors in RADAR Detection Applications

Please enable javascript in your browser.

Sharify Darani, Soroush | 2009

476 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 40350 (05)
University: Sharif University of Technology
Department: Electrical Engineering
Advisor(s): Nayebi, Mohammd Mahdi
Abstract:
In this work, Hough detector has been compared with optimum detectors. We consider 2-D range-time data space, which is constructed by storing the radar returns from different range cells during several successive scans. In this space, target trajectory can be modeled as a straight line for short time intervals, and can be extracted by applying Hough transform transcendentally. In this work, a new detection method based on the likelihood ratio in the data space with Gaussian noise is introduced. ALR and GLR processors have been designed. Then formulas for probability of detection and false alarm rate is calculated, and verified via simulations. A new criterion for estimation quality is introduced and calculated for MAP and Hough processors, too. Finally, detector performance is calculated via power functions and ROC curves, and estimation performance via suitable curves. As the result, the better performance for likelihood ratio processors and MAP estimator than Hough processor has been shown
Keywords:
Radar Detection ; Hough Transform ; Likelihood Ratio ; Range-Time Image ; Avarage Likelihood Ratio (ALR)Test ; Generalized Likelihood Ratio (GLR)Test

No TOC