Precision Target Tracking For Small Extended Objects

Bar‐Shalom, Yaakov; Shertukde, H.M.; Pattipati, Krishna R.

doi:10.1117/12.960340

“…The operations are done in the frequency domain by passing the image through a bank of directional filters each tuned to extract line features of the tracks with certain orientations. Bar-Shalom, et al [9,10] used two different centroid-based measurements to track targets in forward looking IR images. In a more recent paper [11], Shertukde and Bar-Shalom extended this approach by using the joint probabilistic data association (JPDA) in conjunction with a Kalman state estimator.…”

Section: Introductionmentioning

confidence: 99%

“…These include spatio-temporal filtering [3,4], maximum likelihood (ML) estimation [5,12], recursive Kalman filtering [6][7][8][9][10][11] and neural network-based methods [13]. In [4], a 3-D spatio-temporal filtering scheme is developed.…”

Section: Introductionmentioning

confidence: 99%

<title>Neural-network-directed Bayes-decision rule for moving target classification</title>

Azimi-Sadjadi

¹

,

Yu

²

1999

SPIE Proceedings

1

0

View full text Add to dashboard Cite

In this paper, a new neural network directed Bayes decision rule is developed for target classification exploiting the dynamic behavior of the target. The system consists of a feature extractor, a neural network directed conditional probability generator and a novel sequential Bayes classifier. The velocity and curvature sequences extracted from each track are used as the primary features. Similar to hidden Markov model (HMM) scheme, several hidden states are used to train the neural network, the output of which is the conditional probability of occurring the hidden states given the observations. These conditional probabilities are then used as the inputs to the sequential Bayes classifier to make the classification. The classification results are updated recursively whenever a new scan of data is received. Simulation results on multiscan images containing heavy clutter are presented to demonstrate the effectiveness of the proposed methods.

show abstract

“…Several different schemes [3][4][5][6][7][8][9][10][11][12][13] for moving target detection have been developed and applied to IR, radar and sonar imagery data. These include spatio-temporal filtering [3,4], maximum likelihood (ML) estimation [5,12], recursive Kalman filtering [6][7][8][9][10][11] and neural network-based methods [13].…”

Section: Introductionmentioning

confidence: 99%

Neural network directed Bayes decision rule for moving target classification

Yu

¹

,

Azimi-Sadjadi²

2000

IEEE Trans. Aerosp. Electron. Syst.

14

0

View full text Add to dashboard Cite

In this paper, a new neural network directed Bayes decision rule is developed for target classification exploiting the dynamic behavior of the target. The system consists of a feature extractor, a neural network directed conditional probability generator and a novel sequential Bayes classifier. The velocity and curvature sequences extracted from each track are used as the primary features. Similar to hidden Markov model (HMM) scheme, several hidden states are used to train the neural network, the output of which is the conditional probability of occurring the hidden states given the observations. These conditional probabilities are then used as the inputs to the sequential Bayes classifier to make the classification. The classification results are updated recursively whenever a new scan of data is received. Simulation results on multiscan images containing heavy clutter are presented to demonstrate the effectiveness of the proposed methods.

show abstract

Tracking of crossing targets with forward looking infrared imaging sensors

Shertukde

¹

,

Bar‐Shalom²

Proceedings of the 28th IEEE Conference on Decision and Control

3

0

View full text Add to dashboard Cite

Key words:-F o r w a r d looking i n f r a r e d [FLIR] s e n s o r ; Correlation coefficient matrix; colored measurement noise; Joint Probabilistic Data Association [JPDAI Filter. Abstract We p r e s e n t an algorithm f o r t h e tracking of c r o s s i n g t a r g e t s using t h e centroid measurement and the centroid o f f s e t measurement o f the distributed image formed by t h e t a r g e t s . The measurements a r e obtained by a forward looking infrared imaging s e n s o r . Overlap o f t a r g e t images o c c u r s when t h e lines of sight o f t h e t a r g e t s c r o s s . The resulting centroid measurement o f t h e distributed image is a single merged measurement which is a linear combination of t h e c e n t r o i d s o f the individual t a r g e t s under consideration. Also the ensuing image correlation coefficient matrix between two frames i s multimodal and c a r e has to be taken t o properly a s s o c i a t e t h e derived o f f s e t measurements t o a particular t a r g e t . The overlap of the images f o r s e v e r a l sampling times causes a dependence o f t h e s t a t e estimation e r r o r s f o r the two t a r g e t s , which has t o be taken Into account. The Joint Probabilistic Data Association Merged-Measurement Coupled Filter is employed f o r s t a t e estimation which performs filtering in a coupled manner f o r the t a r g e t s with common measurements. Two f i l t e r s a r e examined: one assuming t h e displacement noise white and t h e other one modelling i t c o r r e c t l y as a u t o c o t r e l a t e d . The l a t t e r is shown t o yield substantially b e t t e r performance. The simulation r e s u l t s presented validate the performance predictions of t h e proposed algorithm.

show abstract

Precision Target Tracking For Small Extended Objects

Cited by 3 publications

References 0 publications

<title>Neural-network-directed Bayes-decision rule for moving target classification</title>

<title>Neural-network-directed Bayes-decision rule for moving target classification</title>

Neural network directed Bayes decision rule for moving target classification

Tracking of crossing targets with forward looking infrared imaging sensors

Contact Info

Product

Resources

About