Track-Before-Detect Using Expectation Maximisation by Samuel J. Davey & Han X. Gaetjens

Author:Samuel J. Davey & Han X. Gaetjens
Language: eng
Format: epub
Publisher: Springer Singapore, Singapore

Fig. 7.7Closely spaced non-Gaussian targets

Figure 7.7 shows an example trial where both of the algorithms have established one track. Figure 7.7a shows the sensor frame in the region of the targets. There is a high degree of overlap, but the eye can resolve two targets. Figure 7.7b, c show the output of the frame vetting that is part of the track management described in Chap. 5 for H-PMHT-P and H-PMHT-K respectively. The existing first track in the particle H-PMHT is following the stronger target, but the residual frame clearly shows the response of the second target and the H-PMHT-P was subsequently able to establish a second track. Figure 7.7c shows the output of the frame vetting for H-PMHT-K. Since the algorithm models the target response using a broad Gaussian blob, the vetting suppresses all target information around the location of the extant track and there is almost no sign of the weaker target. The frame shows a dark region where the data has been over-suppressed and a faint outline of the stronger target which has been reduced to the mean background level.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.