Title :
A Bayesian algorithm for object detection in GPR data
Author_Institution :
Inst. for Parallel Process., Bulgarian Acad. of Sci., Sofia
Abstract :
An algorithm for underground object detection in GPR data is presented in this paper. It combines the advantages of Kalman filtering approach, suggested by Carevic (1999) with the robustness of hybrid Bayesian estimation technique. The objective is to increase the reliability of target detection and target-background separation while keeping a small false alarm rate. In addition, sequential change detection (CUSUM) test is studied for the purposes of ground layers segmentation and object recognition.
Keywords :
Bayes methods; Kalman filters; ground penetrating radar; object detection; radar target recognition; Bayesian estimation technique; GPR data; Kalman filtering; ground layer segmentation; ground penetrating radar; object recognition; sequential change detection; target detection; target-background separation; underground object detection; Bayesian methods; Filtering; Ground penetrating radar; Kalman filters; Object detection; Parameter estimation; Robustness; Sequential analysis; State estimation; Technological innovation;
Conference_Titel :
Radar Symposium, 2008 International
Conference_Location :
Wroclaw
Print_ISBN :
978-83-7207-757-8
DOI :
10.1109/IRS.2008.4585776
