激光点云匹配是影响激光SLAM系统精度和效率的关键因素.传统激光SLAM算法无法区分场景结构,且在非结构化场景下由于特征提取不佳而出现性能退化.为此,提出一种联合CPD(coherent point drift)面向复杂场景的自适应激光SLAM算法CPD-LOAM....激光点云匹配是影响激光SLAM系统精度和效率的关键因素.传统激光SLAM算法无法区分场景结构,且在非结构化场景下由于特征提取不佳而出现性能退化.为此,提出一种联合CPD(coherent point drift)面向复杂场景的自适应激光SLAM算法CPD-LOAM.该算法提出一种基于预判和验证相结合的场景结构辨识方法,首先引入场景特征变量对场景结构进行初步判断,然后从几何特征角度通过表面曲率对其进行验证,增强对场景结构辨识的准确性.此外,在非结构化场景下添加CPD算法进行点云预配准,进而利用ICP算法进行再配准,解决该场景下的特征退化问题,从而提高点云配准的精度和效率.实验结果表明,提出的场景特征变量以及表面曲率可以根据设置的阈值有效地区分场景结构,在公开数据集KITTI上的验证结果显示,CPD-LOAM较LOAM算法定位误差降低了84.47%,相较于LeGO-LOAM与LIO-SAM算法定位精度也分别提升了55.88%和30.52%,且具有更高的效率和鲁棒性.展开更多
As image-guided navigation plays an important role in neurosurgery, the spatial registration mapping the pre-operative images with the intra-operative patient position becomes crucial for a high accurate surgical outp...As image-guided navigation plays an important role in neurosurgery, the spatial registration mapping the pre-operative images with the intra-operative patient position becomes crucial for a high accurate surgical output. Conventional landmark-based registration requires expensive and time-consuming logistic support.Surface-based registration is a plausible alternative due to its simplicity and efficacy. In this paper, we propose a comprehensive framework for surface-based registration in neurosurgical navigation, where Kinect is used to automatically acquire patient's facial surface in a real time manner. Coherent point drift(CPD) algorithm is employed to register the facial surface with pre-operative images(e.g., computed tomography(CT) or magnetic resonance imaging(MRI)) using a coarse-to-fine scheme. The spatial registration results of 6 volunteers demonstrate that the proposed framework has potential for clinical use.展开更多
文摘激光点云匹配是影响激光SLAM系统精度和效率的关键因素.传统激光SLAM算法无法区分场景结构,且在非结构化场景下由于特征提取不佳而出现性能退化.为此,提出一种联合CPD(coherent point drift)面向复杂场景的自适应激光SLAM算法CPD-LOAM.该算法提出一种基于预判和验证相结合的场景结构辨识方法,首先引入场景特征变量对场景结构进行初步判断,然后从几何特征角度通过表面曲率对其进行验证,增强对场景结构辨识的准确性.此外,在非结构化场景下添加CPD算法进行点云预配准,进而利用ICP算法进行再配准,解决该场景下的特征退化问题,从而提高点云配准的精度和效率.实验结果表明,提出的场景特征变量以及表面曲率可以根据设置的阈值有效地区分场景结构,在公开数据集KITTI上的验证结果显示,CPD-LOAM较LOAM算法定位误差降低了84.47%,相较于LeGO-LOAM与LIO-SAM算法定位精度也分别提升了55.88%和30.52%,且具有更高的效率和鲁棒性.
基金the National Natural Science Foundation of China(Nos.61190120,61190124 and 61271318)the Biomedical Engineering Fund of Shanghai Jiaotong University(No.YG2012ZD06)
文摘As image-guided navigation plays an important role in neurosurgery, the spatial registration mapping the pre-operative images with the intra-operative patient position becomes crucial for a high accurate surgical output. Conventional landmark-based registration requires expensive and time-consuming logistic support.Surface-based registration is a plausible alternative due to its simplicity and efficacy. In this paper, we propose a comprehensive framework for surface-based registration in neurosurgical navigation, where Kinect is used to automatically acquire patient's facial surface in a real time manner. Coherent point drift(CPD) algorithm is employed to register the facial surface with pre-operative images(e.g., computed tomography(CT) or magnetic resonance imaging(MRI)) using a coarse-to-fine scheme. The spatial registration results of 6 volunteers demonstrate that the proposed framework has potential for clinical use.
