为了在模型参数先验分布知识未知情况下实现基于区域和统计的图像分割,并同时获取更加精确的模型参数,提出了一种结合Voronoi划分技术、最大期望值(Expectation Maximization,EM)和最大边缘概率(Maximizationof the Posterior Marginal,...为了在模型参数先验分布知识未知情况下实现基于区域和统计的图像分割,并同时获取更加精确的模型参数,提出了一种结合Voronoi划分技术、最大期望值(Expectation Maximization,EM)和最大边缘概率(Maximizationof the Posterior Marginal,MPM)算法的图像分割方法。该方法利用Voronoi划分技术将图像域划分为若干子区域,待分割图像中的同质区域可以由一组子区域拟合而成,并假定各同质区域内像素强度服从同一独立的正态分布,从而建立图像模型,然后结合EM/MPM算法进行图像分割和模型参数估计,其中,MPM算法用于实现面向同质区域的图像分割,EM算法用于估计图像模型参数。为了验证提出的图像分割方法,分别对合成图像和真实图像进行了分割实验,并和传统的基于像素的MRF分割结果进行对比,测试结果的定性和定量分析表明了该方法的有效性和准确性。展开更多
The noniterative algorithm of multiscale MRF has much lower computing complexity and better result thanits iterative counterpart of noncausal MRF model, since it has causality property between scales, and such causali...The noniterative algorithm of multiscale MRF has much lower computing complexity and better result thanits iterative counterpart of noncausal MRF model, since it has causality property between scales, and such causality isconsistent with the character of images. Maximizer of the posterior marginals(MPM)algorithm of multiscale MRFmodel is presented for only one image can be obtained in image segmentation. EM algorithm for parameter estimate isalso given. Experiments demonstrate that comparing with iterative ones, the proposed algorithms have the character-istics of greatly reduced computing time and better segmentation results. This is more notable for large images.展开更多
针对船用光纤罗经误差的概率分布不完全符合高斯分布的情况,提出了一种基于高斯混合模型(Gaussian mixture model,GMM)的光纤罗经误差概率分布函数(probability distribution function,PDF)建模方法。该方法使用多个高斯分布的线性叠加...针对船用光纤罗经误差的概率分布不完全符合高斯分布的情况,提出了一种基于高斯混合模型(Gaussian mixture model,GMM)的光纤罗经误差概率分布函数(probability distribution function,PDF)建模方法。该方法使用多个高斯分布的线性叠加来拟合光纤罗经误差的概率分布,并结合一种鲁棒性的期望最大化(expectation maximization,EM)算法来估计GMM中的参数。仿真分析和实测数据验证,相比于使用单一的高斯分布,基于所提方法建立的光纤罗经误差概率分布更加符合该导航设备误差的实际概率分布。展开更多
H-infinity estimator is generally implemented in timevariant state-space models, but it leads to high complexity when the model is used for multiple input multiple output with orthogo- hal frequency division multiplex...H-infinity estimator is generally implemented in timevariant state-space models, but it leads to high complexity when the model is used for multiple input multiple output with orthogo- hal frequency division multiplexing （MIMO-OFDM） systems. Thus, an H-infinity estimator over time-invariant system models is pro- posed, which modifies the Krein space accordingly. In order to avoid the large matrix inversion and multiplication required in each OFDM symbol from different transmit antennas, expectation maximization （EM） is developed to reduce the high computational load. Joint estimation over multiple OFDM symbols is used to resist the high pilot overhead generated by the increasing number of transmit antennas. Finally, the performance of the proposed estimator is enhanced via an angle-domain process. Through performance analysis and simulation experiments, it is indicated that the pro- posed algorithm has a better mean square error （MSE） and bit error rate （BER） performance than the optimal least square （LS） estimator. Joint estimation over multiple OFDM symbols can not only reduce the pilot overhead but also promote the channel performance. What is more, an obvious improvement can be obtained by using the angle-domain filter.展开更多
文摘为了在模型参数先验分布知识未知情况下实现基于区域和统计的图像分割,并同时获取更加精确的模型参数,提出了一种结合Voronoi划分技术、最大期望值(Expectation Maximization,EM)和最大边缘概率(Maximizationof the Posterior Marginal,MPM)算法的图像分割方法。该方法利用Voronoi划分技术将图像域划分为若干子区域,待分割图像中的同质区域可以由一组子区域拟合而成,并假定各同质区域内像素强度服从同一独立的正态分布,从而建立图像模型,然后结合EM/MPM算法进行图像分割和模型参数估计,其中,MPM算法用于实现面向同质区域的图像分割,EM算法用于估计图像模型参数。为了验证提出的图像分割方法,分别对合成图像和真实图像进行了分割实验,并和传统的基于像素的MRF分割结果进行对比,测试结果的定性和定量分析表明了该方法的有效性和准确性。
文摘The noniterative algorithm of multiscale MRF has much lower computing complexity and better result thanits iterative counterpart of noncausal MRF model, since it has causality property between scales, and such causality isconsistent with the character of images. Maximizer of the posterior marginals(MPM)algorithm of multiscale MRFmodel is presented for only one image can be obtained in image segmentation. EM algorithm for parameter estimate isalso given. Experiments demonstrate that comparing with iterative ones, the proposed algorithms have the character-istics of greatly reduced computing time and better segmentation results. This is more notable for large images.
基金supported by the National Natural Science Foundation of China(6087410860904035+2 种基金61004052)the Directive Plan of Science Research from the Bureau of Education of Hebei Province(Z2009105)the Funds of Central Colleges Basic Scientific Operating Expense(N100604004)
文摘H-infinity estimator is generally implemented in timevariant state-space models, but it leads to high complexity when the model is used for multiple input multiple output with orthogo- hal frequency division multiplexing （MIMO-OFDM） systems. Thus, an H-infinity estimator over time-invariant system models is pro- posed, which modifies the Krein space accordingly. In order to avoid the large matrix inversion and multiplication required in each OFDM symbol from different transmit antennas, expectation maximization （EM） is developed to reduce the high computational load. Joint estimation over multiple OFDM symbols is used to resist the high pilot overhead generated by the increasing number of transmit antennas. Finally, the performance of the proposed estimator is enhanced via an angle-domain process. Through performance analysis and simulation experiments, it is indicated that the pro- posed algorithm has a better mean square error （MSE） and bit error rate （BER） performance than the optimal least square （LS） estimator. Joint estimation over multiple OFDM symbols can not only reduce the pilot overhead but also promote the channel performance. What is more, an obvious improvement can be obtained by using the angle-domain filter.
