Estimating the Parameters of General Frequency Modulated Signals

Luginbuhl, Tod; Willett, Peter

doi:10.1109/tsp.2003.820080

Cited by 35 publications

(21 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The updated target state estimate is found by maximizing (9) where H = i h i x X k ;π k , which is generally very close to unity, and B i is the state-space extent of pixel i.…”

Section: H-pmhtmentioning

confidence: 99%

“…However unlike the Gaussian case, there exists no analytic way of reducing the auxiliary function (9) to an equivalent point measurement problem with the same likelihood. Nevertheless, the EM procedure can still be employed to return estimates of the mixing proportions and target states using non-linear state estimation methods.…”

Section: H-pmhtmentioning

confidence: 99%

“…The Histogram Probabilistic Multi-Hypothesis Tracker (H-PMHT) is an example of another unique algorithm developed to address the TkBD problem [9], [17]. Its derivation is based on the generation of an artificial histogram by quantizing the sensor data and then the application of ExpectationMaximization (EM) [8] based mixture modeling to describe the underlying data sources.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Track-Before-Detect Using Histogram PMHT and Dynamic Programming

Davey

2012

2012 International Conference on Digital Image Computing Techniques and Applications (DICTA)

View full text Add to dashboard Cite

The Histogram Probabilistic Multi-Hypothesis Tracker (H-PMHT) is a parametric mixture-fitting approach to the Track-before-detect (TkBD) problem. It has been shown to give performance close to numerical approximations of the optimal Bayesian filter at a fraction of the computation cost. This paper will consider an implementation of the H-PMHT for non-linear non-Gaussian TkBD problems using a dynamic programming fixed-grid approximation through application of the Viterbi algorithm. This alternate H-PMHT implementation is compared with Kalman Filter and Particle Filter H-PMHT implementations via simulated single target scenarios.

show abstract

“…The updated target state estimate is found by maximizing (9) where H = i h i x X k ;π k , which is generally very close to unity, and B i is the state-space extent of pixel i.…”

Section: H-pmhtmentioning

confidence: 99%

Section: H-pmhtmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Track-Before-Detect Using Histogram PMHT and Dynamic Programming

Davey

2012

2012 International Conference on Digital Image Computing Techniques and Applications (DICTA)

View full text Add to dashboard Cite

show abstract

“…Using the EM algorithm for truncated two-dimensional Gaussian mixture models described in [11]- [13], the parame-…”

Section: A Acf Model Parameter Estimationmentioning

confidence: 99%

Autocorrelation features for synthetic aperture sonar image seabed segmentation

Cobb

Prı́ncipe

2011

2011 IEEE International Conference on Systems, Man, and Cybernetics

View full text Add to dashboard Cite

Abstract-High-resolution synthetic aperture sonar (SAS) systems yield richly detailed images of seabed environments. Algorithms that automatically segment and label seabed textures such as coral, sea grass, sand ripple, and mud, require suitable features that discriminate between the texture classes. Here we present a robust, parameterized SAS image texture model based on the autocorrelation function (ACF) of the intensity image. This ACF texture model has been shown to accurately model first-and second-order statistical features of various seabed environments. An unsupervised multi-class k-means segmentation algorithm that uses the features derived from the ACF model is employed to label rock and ripple textures from a set of textured SAS images. The results of the segmentation are compared against the performance of the segmentation approach using biorthogonal wavelets and Haralick features. In the described experiments, the ACF model features are shown to produce better segmentations than the features based on wavelet coefficients and Haralick features for classifiers of low complexity.

show abstract

“…The standard PMHT restricts the prior to be either constant, or time independent. Existing PMHT research does not address the assignment prior; effort has instead focussed on the target dynamics model [4,8,6,9], more sophisticated measurement models [10][11][12], and matters of practical significance for realistic implementations [7,13,14]. A model is proposed here which allows the assignment prior to be a randomly evolving quantity.…”

Section: Introductionmentioning

confidence: 99%