Frequency line tracking using hidden Markov models

Streit, Roy L.; Barrett, Ross

doi:10.1109/29.52700

Cited by 137 publications

(51 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, there are many examples of HMMs being employed in detection, tracking and identification using the observations obtained by remote sensors such as radar and sonar. For instance in [9] the problem of frequency line tracking is tackled by employing HMMs and in [10] the problem is extended to multiple frequency line tracking with ambiguous detections in which the problem was formulated in terms of HMM to produce MAP track estimates via the Viterbi algorithm. Tracking and target motion analysis (TMA) were performed in [11] by discretizing the target states in a grid of possible positions and speeds where the target state evolutions were assumed to be stochastic.…”

Section: Introductionmentioning

confidence: 99%

Radar Target Detection Using Hidden Markov Models

Tuğaç¹,

Efe²

2012

PIER B

View full text Add to dashboard Cite

Abstract-Standard radar detection process requires that the sensor output is compared to a predetermined threshold. The threshold is selected based on a-priori knowledge available and/or certain assumptions. However, any knowledge and/or assumptions become inadequate due to the presence of multiple targets with varying signal return and usually non stationary background. Thus, any fixed predefined threshold may result in either increased false alarm rate or increased track loss. Even approaches where the threshold is adaptively varied will not perform well in situations when the signal return from the target of interest is too low compared to the average level of the background. Track-before-detect techniques eliminate the need for a detection threshold and provide detecting and tracking targets with lower signal-to-noise ratios than standard methods. However, although track-before-detect techniques eliminate the need for detection threshold at sensor's signal processing stage, they often use tuning thresholds at the output of the filtering stage. This paper presents a Hidden Markov Model based target detection method that avoids any thresholding at any stage of the detection process. Moreover, since the proposed Hidden Markov Model method is based on the target motion models, the output of the detection process can easily be employed for maneuvering target tracking.

show abstract

Section: Introductionmentioning

confidence: 99%

Radar Target Detection Using Hidden Markov Models

Tuğaç¹,

Efe²

2012

PIER B

View full text Add to dashboard Cite

show abstract

“…[14], [4]), but their application to localize sound sources has not yet been explored. Other related works, that however do not concern sound aspects, are [1], [2], [5].…”

Section: Introductionmentioning

confidence: 99%

Tracking sound sources by means of HMM

Rodà

Micheloni

2011

2011 8th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS)

View full text Add to dashboard Cite

Video-based surveillance systems may benefit from the integration with microphone arrays for the localization of sound events. Applying the sound localization techniques to the surveillance of large areas requires addressing some open issues, such as the non uniform resolution of the microphones-based localization systems. This paper presents a new method for tracking moving sound events based on an Hidden Markov Model (HMM), which exploits a priori information derived from medium and longterm observations of the monitored area. The results obtained with simulated trajectories show that the HMMbased tracker is able to significantly reduce the localization error. Applications can be found in surveillance systems for large areas, such as square, streets, or parking lots, where it is of interest the monitoring of moving vehicles and people.

show abstract

“…Similarly, Johansson and White [2] apply the notch filter concept to time variant auto regressive data modeling, leading to two types of time-variant notch filters that are applied to signals with low signal-to-noise ratios. In [3], Streit and Barrett formulate the problem of frequency line tracking as an optimal path through fixed frequency states, determined by a Hidden Markov Model which implies a high noise robustness at the cost of a coarse frequency resolution (prior chosen states). Witte et al [4] use an active contour model approach from image processing, the Snake algorithm, to extract a time frequency line.…”

Section: Introductionmentioning

confidence: 99%

Frequency Tracking with Spline Based Chirp Atoms

Wacker

Galicki

Witte

2010

Lecture Notes in Computer Science

View full text Add to dashboard Cite

The tracking of prominent oscillations with time-varying frequencies is a common task in many signal processing applications. Therefore, efficient methods are needed to meet precision and run-time requirements. We propose two methods that extract the energy of the time frequency plane along a cubic spline trajectory. To raise efficiency, a sparse sampling method with dynamically shaped atoms is developed for the one method in contrast to standard Gabor atoms for the other. Numerical experiments using both synthetic and real-life data are carried out to show that dynamic atoms can significantly outperform classical Gabor formulations for this task.

show abstract

Frequency line tracking using hidden Markov models

Cited by 137 publications

References 11 publications

Radar Target Detection Using Hidden Markov Models

Radar Target Detection Using Hidden Markov Models

Tracking sound sources by means of HMM

Frequency Tracking with Spline Based Chirp Atoms

Contact Info

Product

Resources

About