Multirate interacting multiple model particle filter for terrain-based ground target tracking

Liu, Hong; Cui, Ningzhou; Bakich, Michael; Layne, J.R.

doi:10.1049/ip-cta:20050047

Cited by 28 publications

(13 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The major difference from the traditional wavelet transform in Eqs. (11) and (12) is that x i 0 , i = 1, ..., N are random samples, not a predefined and evenly-spaced sequence. To overcome the randomness, we sample all N0 particles first; then filtering is applied, but the samples are still unevenly spaced.…”

Section: Haar Waveletsmentioning

confidence: 99%

“…On the other hand, applying multiresolutional processing in a spatial domain in designing a particle filter results in a multires particle filter. In our previous work, we have demonstrated significant computational savings by using a multirate multiple model particle filter [11]. This paper is focused on computational savings with a multires particle filter.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A spatial domain multiresolutional particle filter

Liu

Xue

2007

2007 Mediterranean Conference on Control &Amp; Automation

View full text Add to dashboard Cite

Particle filters have been proven to be very effective for nonlinear/non-Gaussian filtering. However, the most notorious disadvantage of a particle filter is its formidable computational complexity, since hundreds (even thousands) of particles are usually needed to achieve a required approximation accuracy. It has also been proven that one of the techniques of truly solving a computational problem is multiresolutional processing, both in temporal and spatial domains. Therefore, in this paper we propose a multiresolutional particle filter in the spatial domain using thresholded wavelets to reduce significantly the number of particles, meanwhile maintaining the full strength of a particle filter.

show abstract

Section: Haar Waveletsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A spatial domain multiresolutional particle filter

Liu

Xue

2007

2007 Mediterranean Conference on Control &Amp; Automation

View full text Add to dashboard Cite

show abstract

“…Hence, multiple model (MM) based approaches are often used for maneuvering target tracking [3]. To have both the merits of MM-based approaches and PFs, several researchers tried to combine them to solve the maneuvering target tracking problems [4][5][6][7][8][9]. Generally, MMPFs have one non-maneuver mode and several maneuver modes with different degree of acceleration, and assume that the Markov chain transition probabilities are stationary and known.…”

Section: Introductionmentioning

confidence: 99%

Easy-hardware-implementation MMPF for Maneuvering Target Tracking: Algorithm and Architecture

Hong

Shi

Chen

2010

J Sign Process Syst

View full text Add to dashboard Cite

In this paper, we present an easy-hardwareimplementation multiple model particle filter (MMPF) for maneuvering target tracking. In the proposed filter, the sampling importance resampling (SIR) filter typically used for nonlinear and/or non-Gaussian application is extended to incorporating multiple models that are composed of a constant velocity (CV) model and a "current" statistical (CS) model, and the Independent Metropolis Hasting (IMH) sampler is utilized for the resampling unit in each model. Compared with the bootstrap MMPF, the proposed MMPF requires no knowledge of models and model transition probabilities for different maneuvering motions, and keeps a constant number of particles per model at all times. This allows a regular pipelined hardware structure and can be implemented in hardware easily. Furthermore, using the IMH sampler for the resampling unit avoids the bottleneck introduced by the traditional systematic resampler and reduces the latency of the whole implementation. Simulation results indicate that the proposed filter has approximately equal tracking performance with the bootstrap MMPF. Hardware architecture of the IMH sampler and its corresponding sample unit are presented, and a parallel architecture consisting of CV model processing element (PE), CS model PE and a central unit (CU) is described. The proposed architecture is evaluated on a Xilinx Virtex-II Pro FPGA platform for a maneuvering target tracking application and the results show many advantages of the proposed MMPF over existing approaches in terms of efficiency, lower latency, and easy hardware implementation.

show abstract

“…When the number of observation is large, multiple target tracking algorithms such as [11] are introduced to handle data association for objects with similar appearance. The IMM-EKF [12] is introduced to fuse multiple models, which provides an open framework to cover cases such as the object performing multiple behaviors [13], [14], observations come from different types of sensors [15], and tracking with different sampling periods [16]. A more relevant work to this paper is proposed in [17], where shape models and behavior models are combined to simultaneously track and analysis bee dances, using particle filter.…”

mentioning

confidence: 99%

Automatic Bird Species Filtering Using a Multimodel Approach

Song

2015

IEEE Trans. Automat. Sci. Eng.

View full text Add to dashboard Cite

We report a filtering algorithm for bird species detection using videos captured by uncalibrated moving cameras, a typical characteristic of crowd sourced videos. The algorithm tracks both body and wing motion dimensions of a flying bird to form signatures for species filtering. In the body model, we consider both cases when background motion introduced by the camera can and cannot be directly recognized using key point matching. We are also able to recover intrinsic camera parameters in the body motion tracking. In the wing model, we consider both periodic wing flapping and gliding motion patterns. These models are combined to form a multimodel framework. We have tested the algorithm and compared its performance with single model approaches in physical experiments. Results show that the new algorithm significantly reduces the false positive rate, while maintaining a low false negative rate. The area under ROC curve is 92.86%. Note to Practitioners-This paper attempts to bring automation to field biologists. This is relatively new area as videos are often captured under natural lightings by non-experts. Hence, new challenges arrive as the demand for analyzing these large amount of video data increases. We propose a species detection algorithm for birds by considering factors such as unknown camera intrinsics and motion, variable lighting conditions, and different backgroundtypes. This algorithm have been tested with crowd sourced video data with different species of birds. The results meet the design goals and it proves that this algorithm can be an effective tool to assist ornithologists.Index Terms-Bird species recognition, interactive multimodel filtering, nature observation.

show abstract

Multirate interacting multiple model particle filter for terrain-based ground target tracking

Cited by 28 publications

References 12 publications

A spatial domain multiresolutional particle filter

A spatial domain multiresolutional particle filter

Easy-hardware-implementation MMPF for Maneuvering Target Tracking: Algorithm and Architecture

Automatic Bird Species Filtering Using a Multimodel Approach

Contact Info

Product

Resources

About