针对海上低纹理、大视差图像拼接过程中由于误对齐产生的伪影问题,提出一种基于点线特征配准和最佳接缝线融合的几何结构保视差图像拼接算法。在传统的基于点特征求解单应变换的基础上引入线段特征,并将潜在的共面局部线段合并为全局线...针对海上低纹理、大视差图像拼接过程中由于误对齐产生的伪影问题,提出一种基于点线特征配准和最佳接缝线融合的几何结构保视差图像拼接算法。在传统的基于点特征求解单应变换的基础上引入线段特征,并将潜在的共面局部线段合并为全局线段,为接缝线融合提供精准对齐的条件;在图像融合过程中,利用tanh度量的颜色差异和梯度差异以及引入显著性检测权重来设计接缝切割方法的能量函数,引导最佳接缝线避开图像中显著海上结构,从而确保结构边缘的连续性;使用SIFT flow方法校正拼接缝上的错位像素,实现几何结构准确的海上图像拼接。在20对不同场景数据上的拼接实验表明,与基准方法相比,所提算法的基于SSIM的接缝质量误差平均降低了44.6%,最多降低66.7%,基于ZNCC(zero mean normalized cross-correlation)的接缝质量误差平均降低了24.7%,最多降低了51.6%,能够有效地避免伪影,从而得到观感自然的拼接结果,满足无人船航行过程中对宽视场的需求。展开更多
This paper introduces the research progress on the cognition flow in the National Basic Research Program of China(973 Program)under Grant No. 2009CB320400 "Research on Basic Theories and Key Technologies of Cogni...This paper introduces the research progress on the cognition flow in the National Basic Research Program of China(973 Program)under Grant No. 2009CB320400 "Research on Basic Theories and Key Technologies of Cognitive Radio Networks". We present the motivation behind the proposal of the concept of cognition flow, provide the definition, discuss the features, behaviours, representations, mathematical models, and the functions of cognition flow in Cognitive Radio Networks(CRNs). We also analyse how the cognition flow promotes the convergence of heterogeneous networks.Our group also constructed the test platform to verify the usefulness of cognition flow. The results that were simulated by computers and tested on the platform both confirm that cognition flow can realise efficient interaction of cognitive information among heterogeneous networks in CRNs, which contributes to the seamless convergence of heterogeneous networks,and significantly improves the spectrum efficiency and users' Quality of Experience(QoE).展开更多
文摘针对海上低纹理、大视差图像拼接过程中由于误对齐产生的伪影问题,提出一种基于点线特征配准和最佳接缝线融合的几何结构保视差图像拼接算法。在传统的基于点特征求解单应变换的基础上引入线段特征,并将潜在的共面局部线段合并为全局线段,为接缝线融合提供精准对齐的条件;在图像融合过程中,利用tanh度量的颜色差异和梯度差异以及引入显著性检测权重来设计接缝切割方法的能量函数,引导最佳接缝线避开图像中显著海上结构,从而确保结构边缘的连续性;使用SIFT flow方法校正拼接缝上的错位像素,实现几何结构准确的海上图像拼接。在20对不同场景数据上的拼接实验表明,与基准方法相比,所提算法的基于SSIM的接缝质量误差平均降低了44.6%,最多降低66.7%,基于ZNCC(zero mean normalized cross-correlation)的接缝质量误差平均降低了24.7%,最多降低了51.6%,能够有效地避免伪影,从而得到观感自然的拼接结果,满足无人船航行过程中对宽视场的需求。
基金supported by the National Basic Research Program of China(973 Program)under Grant No.2009CB320400the National Natural Science Foundation of China under Grants No.61101117,No.61171099+2 种基金the National Key Scientific and Technological Project of China under Grant No,2012ZX03003-007the Natural Science Foundation of Jiangxi under Grant No.20132BAB201018the Fundamental Research Funds for the Central Universities under Grant No.BUPT2012RC0112
文摘This paper introduces the research progress on the cognition flow in the National Basic Research Program of China(973 Program)under Grant No. 2009CB320400 "Research on Basic Theories and Key Technologies of Cognitive Radio Networks". We present the motivation behind the proposal of the concept of cognition flow, provide the definition, discuss the features, behaviours, representations, mathematical models, and the functions of cognition flow in Cognitive Radio Networks(CRNs). We also analyse how the cognition flow promotes the convergence of heterogeneous networks.Our group also constructed the test platform to verify the usefulness of cognition flow. The results that were simulated by computers and tested on the platform both confirm that cognition flow can realise efficient interaction of cognitive information among heterogeneous networks in CRNs, which contributes to the seamless convergence of heterogeneous networks,and significantly improves the spectrum efficiency and users' Quality of Experience(QoE).
