开源工具：实现了二十来种视频前景提取的算法（bgslibrary）
BGSLibrary ：A Background Subtraction Library，实现了二十来种视频前景提取的算法。不一定每个都最优，但是可以做一些简单的对比。
The BGSLibrary was developed by Andrews Sobral and provides a C++ framework to perform background subtraction (BGS). The code works either on Windows or on Linux. Currently the library offers 31 BGS algorithms. A large amount of algorithms were provided by several authors. The source code is available under GNU GPL v3 license, the library is free and open source. Any user can be download latest project source code using SVN client. In Windows, a demo project for Visual Studio 2010 is provided. An executable version of BGSLibrary is available for Windows 32 bits and 64 bits. For Linux users, a Makefile can be used to compile all files and generate an executable example. Table 1 lists some of algorithms available in BGSLibrary. The algorithms are classified by their similarities.
地址：https://code.google.com/p/bgslibrary/
Basic methods, mean and variance overtime:
(StaticFrameDifferenceBGS) Static FrameDifference
(FrameDifferenceBGS) FrameDifference
(WeightedMovingMeanBGS) WeightedMoving Mean
(WeightedMovingVarianceBGS) WeightedMoving Variance
(AdaptiveBackgroundLearning) AdaptiveBackground Learning
1 (DPMeanBGS) Temporal Mean
1 (DPAdaptiveMedianBGS) AdaptiveMedianof McFarlaneand Schofield (1995) paper link
1 (DPPratiMediodBGS) TemporalMedianof Cucchiaraet al (2003) and Calderara et al (2006) paper link1paperlink2 paper link3
Fuzzy based methods:
2 (FuzzySugenoIntegral) Fuzzy SugenoIntegral(with Adaptive-SelectiveUpdate) of Hongxun Zhang and De Xu (2006) paper link
2 (FuzzyChoquetIntegral) Fuzzy ChoquetIntegral(withAdaptive-Selective Update) of Baf et al (2008) paper link
3 (LBFuzzyGaussian) FuzzyGaussianof Sigari etal (2008) paper link
Statistical methods using one gaussian:
1 (DPWrenGABGS) GaussianAverageof Wren(1997) paper link
3 (LBSimpleGaussian) SimpleGaussianof Benezethet al (2008) paper link
Statistical methods using multiplegaussians:
1 (DPGrimsonGMMBGS) GaussianMixture Modelof Staufferand Grimson (1999) paper link
0 (MixtureOfGaussianV1BGS) GaussianMixture ModelofKadewTraKuPong and Bowden (2001)paper link
0 (MixtureOfGaussianV2BGS) GaussianMixture Modelof Zivkovic(2004) paper link1paper link2
1 (DPZivkovicAGMMBGS) GaussianMixture Modelof Zivkovic(2004) paper link1paper link2
3 (LBMixtureOfGaussians) GaussianMixture Modelof Baf et al(2008) paper link
Type-2 Fuzzy based methods:
2 (T2FGMM_UM) Type-2 FuzzyGMM-UMof Baf et al(2008) paper link
2 (T2FGMM_UV) Type-2 FuzzyGMM-UVof Baf et al(2008) paper link
2 (T2FMRF_UM) Type-2 FuzzyGMM-UM with MRFof Zhao et al(2012) paper link1paper link2
2 (T2FMRF_UV) Type-2 FuzzyGMM-UV with MRFof Zhao et al(2012) paper link1paper link2
Statistical methods using color andtexture features:
1 (DPTextureBGS) Texture BGSof Heikkila et al.(2006) paper link
4 (MultiLayerBGS) Multi-LayerBGSof Jian Yaoand Jean-Marc Odobez (2007) paper link
Non-parametric method:
5 (PixelBasedAdaptiveSegmenter) Pixel-BasedAdaptive Segmenter (PBAS)of Hofmann etal (2012) paperlink
0 (GMG) GMG of Godbehere et al(2012)paper link
6 (VuMeter) VuMeter of Goyat et al (2006) paper link
7 (KDE) KDE of Elgammal et al (2000)paper link
Methods based on eigen features:
1 (DPEigenbackgroundBGS) Eigenbackground/ SL-PCAof Oliver etal (2000) paper link
Neural and neuro-fuzzy methods:
3 (LBAdaptiveSOM) Adaptive SOMof Maddalena andPetrosino (2008) paper link
3 (LBFuzzyAdaptiveSOM) FuzzyAdaptive SOMof Maddalenaand Petrosino (2010) paper link
| Legend: |
| 0 native from OpenCV |
| 1adapted from Donovan Parks |
| 2adapted from Thierry Bouwmans and Zhenjie Zhao |
| 3adapted from Laurence Bender |
| 4adapted from Jian Yao and Jean-Marc Odobez |
| 5adapted from Martin Hofmann, Philipp Tiefenbacher and Gerhard Rigoll |
| 6adapted from Lionel Robinault and Antoine Vacavant |
| 7adapted from Ahmed Elgammal |
------------------------------------------------------------------------
版权声明：
转自：http://www.sigvc.org/bbs/thread-993-1-1.html