A detailed study was carried out to find optimal seam-lines for mosaicking of images acquired by an airborne light detection and ranging (LiDAR) system.High ground objects labeled as obstacles can be identified by del...A detailed study was carried out to find optimal seam-lines for mosaicking of images acquired by an airborne light detection and ranging (LiDAR) system.High ground objects labeled as obstacles can be identified by delineating black holes from filtered point clouds obtained by filtering the raw laser scanning dataset.An innovative A algorithm is proposed that can automatically make the seam-lines keep away from these obstacles in the registered images.This method can intelligently optimize the selection of seam-lines and improve the quality of orthophotos.A simulated grid image was first used to analyze the effect of different heuristic functions on path planning.Three subsets of LiDAR data from Xi'an,Dunhuang,and Changyang in Northwest China were obtained.A quantitative method including pixel intensity,hue,and texture was used.With our proposed method,9.4%,8.7%,and 9.8% improvements were achieved in Dunhuang,Xi'an,and Changyang,respectively.展开更多
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
基金supported by the National Basic Research Program (973) of China (No.2009CB724007)the National High-Tech R & D Program (863) of China (No.2006AA12Z101)
文摘A detailed study was carried out to find optimal seam-lines for mosaicking of images acquired by an airborne light detection and ranging (LiDAR) system.High ground objects labeled as obstacles can be identified by delineating black holes from filtered point clouds obtained by filtering the raw laser scanning dataset.An innovative A algorithm is proposed that can automatically make the seam-lines keep away from these obstacles in the registered images.This method can intelligently optimize the selection of seam-lines and improve the quality of orthophotos.A simulated grid image was first used to analyze the effect of different heuristic functions on path planning.Three subsets of LiDAR data from Xi'an,Dunhuang,and Changyang in Northwest China were obtained.A quantitative method including pixel intensity,hue,and texture was used.With our proposed method,9.4%,8.7%,and 9.8% improvements were achieved in Dunhuang,Xi'an,and Changyang,respectively.
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。
