Background subtraction models based on mixture of Gaussians have been extensively used for detecting objects in motion in a wide variety of computer vision applications. However, background subtraction modeling is sti...
详细信息
Background subtraction models based on mixture of Gaussians have been extensively used for detecting objects in motion in a wide variety of computer vision applications. However, background subtraction modeling is still an open problem particularly in video scenes with drastic illumination changes and dynamic backgrounds (complex backgrounds). The purpose of the present work is focused on increasing the robustness of background subtraction models to complex environments. For this, we proposed the following enhancements: a) redefine the model distribution parameters involved in the detection of moving objects (distribution weight, mean and variance), b) improve pixel classification (background/foreground) and variable update mechanism by a new time-space dependent learning-rate parameter, and c) replace the pixel-based modeling currently used in the literature by a new space-time region-based model that eliminates the noise effect caused by drastic changes in illumination. Our proposed scheme can be implemented on any state of the art background subtraction scheme based on mixture of Gaussians to improve its resilient to complex backgrounds. Experimental results show excellent noise removal and object motion detection properties under complex environments.
Identification of the Switched Markov Autoregressive exogenous (ARX) systems is considered in this paper. With a Markov chain model governing the evolution of the hidden switching state, a Switched Markov ARX System (...
详细信息
Identification of the Switched Markov Autoregressive exogenous (ARX) systems is considered in this paper. With a Markov chain model governing the evolution of the hidden switching state, a Switched Markov ARX System (SMARX) is formulated and a solution strategy is proposed. The expectation-maximization (EM) algorithm is employed in the identification of the SMARX systems in which both a Hidden Markov Model (HMM) for the discrete-valued switching dynamics and local ARX models for continuous dynamics are estimated. Through the comparison between the proposed method and previous switched ARX system identification methods, it is shown that by modeling both the switching and continuous dynamics, the accuracy of the identification results can, to various extent, be improved. (C) 2011 Elsevier Ltd. All rights reserved.
Data-driven soft sensors have been applied extensively in process industry for process monitoring and control. Linear soft sensors, which are only valid within a relatively small operating envelope, are considered to ...
详细信息
Data-driven soft sensors have been applied extensively in process industry for process monitoring and control. Linear soft sensors, which are only valid within a relatively small operating envelope, are considered to be insufficient in practice when the processes transit among several operating modes. Moreover, owing to a variety of causes such as malfunction of sensors, multiple rate sampling scheme for different process variables, etc., missing data problem is commonly experienced in process industry. In this paper, soft sensor development with irregular/missing output data is considered and a multiple model based linear parameter varying (LPV) modeling scheme is proposed for handling nonlinearity. The efficiency of the proposed algorithm is demonstrated through several numerical simulation examples as well as a pilot-scale experiment. It is shown through the comparison with the traditional missing data treatment methods in terms of the parameter estimation accuracy that the developed soft sensors enjoy improved performance by employing the expectation-maximization (EM) algorithm in handling the missing process data and model switching problem. (C) 2011 Elsevier Ltd. All rights reserved.
This paper is concerned with the identification of a nonlinear process which operates over several working points with consideration of transition dynamics between the working points. Operating point changes due to ec...
详细信息
This paper is concerned with the identification of a nonlinear process which operates over several working points with consideration of transition dynamics between the working points. Operating point changes due to economic considerations (e.g. grade change in polymer plants) or working environment changes (e.g. feed raw materials property change) are commonly experienced in process industry. These transitions among different operating conditions excite the inherent nonlinearity of the chemical process and pose significant challenges for process modeling. To circumvent the difficulties, we propose a probability-based identification method in which a linear parameter varying (LPV) model is built using process input-output data. Without knowing the local model dynamics a priori, only excitation signals around each operating point are required to identify linear models of the local dynamics, and then the local models are synthesized with transition data to construct a global LPV model. Simulated numerical examples as well as an experiment performed on a pilot-scale heated tank are employed to demonstrate the effectiveness of the proposed method. (C) 2010 Elsevier Ltd. All rights reserved.
In this paper, a real time method for detecting multiple dim targets in deep space background is presented and special attention is paid to occlusion handling. We matched the stars in tow continuous images to get thei...
详细信息
ISBN:
(纸本)9780819488374
In this paper, a real time method for detecting multiple dim targets in deep space background is presented and special attention is paid to occlusion handling. We matched the stars in tow continuous images to get their speed at first and found moving target pairs through speed in both images, a kalman filter whose equation was updated by the centroid was adopted to track the target. The star's area was used to judge occlusion, a two Gaussian mixture model was build using the pixels' gray value of fusing region and we used the predicted value which the kalman filter given to detect the target. The model's parameters were estimated using the expectation-maximization method and applied to separate the target and the star as well as computing the precise centroid. Extensive experiments on real images sequences show that the proposed approach could effectively meet the requirements of the real-time detection with a low false alarm rate and a high detection probability, simulation results show that it can also create a accuracy centroid when occlusion happens.
A methodology for the identification of nonlinear models using constrained particle filters under the scheme of the expectation-maximization (EM) algorithm is presented in this paper. Missing or irregularly sampled ob...
详细信息
ISBN:
(纸本)9781457700811
A methodology for the identification of nonlinear models using constrained particle filters under the scheme of the expectation-maximization (EM) algorithm is presented in this paper. Missing or irregularly sampled observations are commonplace in the chemical industry. In order to circumvent the difficulties rendered by largely incomplete data set, an improved EM based algorithm, which uses the expected value of the log-likelihood function including the missing observations, is developed. Constrained particle filters are adopted to solve the expected log-likelihood function in the EM algorithm. The efficiency of the proposed method in handling missing data is illustrated through numerical examples and validated through experiments.
A key issue in the development and deployment of model-based automatic target recognition (ATR) systems is the generation of target models to populate the ATR database. Model generation is typically a formidable task,...
详细信息
A key issue in the development and deployment of model-based automatic target recognition (ATR) systems is the generation of target models to populate the ATR database. Model generation is typically a formidable task, often requiring detailed descriptions of targets in the form of blueprints or CAD models. We propose a method for generating a 3-D target model directly from multiple SAR images of a target obtained at arbitrary viewing angles. This 3-D model is a parameterized description of the target in terms of its component reflector primitives. We pose the model generation problem as a parametric estimation problem based on information extracted from the SAR images. We accomplish this parametric estimation in the context of data association using the expectation-maximization (EM) method. Our model generation technique operates without supervision and adaptively selects the model order. Although we develop our method in the context of a specific data extraction technique and target parameterization scheme, our underlying framework is general enough to accommodate different choices. We present results demonstrating the utility of our method. (C) 2002 society of Photo-Optical Instrumentation Engineers.
Algorithms to treat the maximum likelihood (ML) estimation problem in array localization signal processing are reviewed, including the alternating projection method, the iterative quadratic maximum likelihood method a...
详细信息
Algorithms to treat the maximum likelihood (ML) estimation problem in array localization signal processing are reviewed, including the alternating projection method, the iterative quadratic maximum likelihood method and the expectation-maximization method. The relationship of ML estimators and the MUSIC algorithm is presented. The Cramer-Rao bounds for the deterministic and stochastic models in array localization are summarized. Finally, the problem of the estimation of the number of sources is discussed.
暂无评论