A novel and effective approach to global motion estimation and moving object extraction is proposed. First, the translational motion model is used because of the fact that complex motion can be decomposed as a sum of ...A novel and effective approach to global motion estimation and moving object extraction is proposed. First, the translational motion model is used because of the fact that complex motion can be decomposed as a sum of translational components. Then in this application, the edge gray horizontal and vertical projections are used as the block matching feature for the motion vectors estimation. The proposed algorithm reduces the motion estimation computations by calculating the onedimensional vectors rather than the two-dimensional ones. Once the global motion is robustly estimated, relatively stationary background can be almost completely eliminated through the inter-frame difference method. To achieve an accurate object extraction result, the higher-order statistics （HOS） algorithm is used to discriminate backgrounds and moving objects. Experimental results validate that the proposed method is an effective way for global motion estimation and object extraction.展开更多
In this work, we established a converse duality theorem for higher-order Mond-Weir type multiob- jective programming involving cones. This fills some gap in recently work of Kim et al. [Kim D S, Kang H S, Lee Y J, et ...In this work, we established a converse duality theorem for higher-order Mond-Weir type multiob- jective programming involving cones. This fills some gap in recently work of Kim et al. [Kim D S, Kang H S, Lee Y J, et al. Higher order duality in inultiobjective programming with cone constraints. Optimization, 2010, 59： 29-43].展开更多
This paper extends the class of generalized type I functions introduced by Aghezzaf and Hachimi（2000） to the context of higher-order case and formulate a number of higher-order duals to a non-differentiable multi-ob...This paper extends the class of generalized type I functions introduced by Aghezzaf and Hachimi（2000） to the context of higher-order case and formulate a number of higher-order duals to a non-differentiable multi-objective programming problem and establishes higher-order duality results under the higher-order generalized type I functions introduced in the present paper, A special case that appears repeatedly in the literature is that the support function is the square root of a positive semi-definite quadratic form. This and other special cases can be readily generated from these results.展开更多
基金The National Natural Science Foundation of China(No.60574006)
文摘A novel and effective approach to global motion estimation and moving object extraction is proposed. First, the translational motion model is used because of the fact that complex motion can be decomposed as a sum of translational components. Then in this application, the edge gray horizontal and vertical projections are used as the block matching feature for the motion vectors estimation. The proposed algorithm reduces the motion estimation computations by calculating the onedimensional vectors rather than the two-dimensional ones. Once the global motion is robustly estimated, relatively stationary background can be almost completely eliminated through the inter-frame difference method. To achieve an accurate object extraction result, the higher-order statistics （HOS） algorithm is used to discriminate backgrounds and moving objects. Experimental results validate that the proposed method is an effective way for global motion estimation and object extraction.
基金supported by National Natural Science Foundation of China(Grant Nos.10831009 and 11271391)the Natural Science Foundation of Chongqing(Grant No.CSTC2011BA0030)
文摘In this work, we established a converse duality theorem for higher-order Mond-Weir type multiob- jective programming involving cones. This fills some gap in recently work of Kim et al. [Kim D S, Kang H S, Lee Y J, et al. Higher order duality in inultiobjective programming with cone constraints. Optimization, 2010, 59： 29-43].
文摘This paper extends the class of generalized type I functions introduced by Aghezzaf and Hachimi（2000） to the context of higher-order case and formulate a number of higher-order duals to a non-differentiable multi-objective programming problem and establishes higher-order duality results under the higher-order generalized type I functions introduced in the present paper, A special case that appears repeatedly in the literature is that the support function is the square root of a positive semi-definite quadratic form. This and other special cases can be readily generated from these results.
