A retrospective analysis of the prognosis of Graves disease was carried on 224 cases, whose treatment has been stopped for 6 months to more than 10 years. The patients were divided into 3 groups. Group 1. 92 cases, re...A retrospective analysis of the prognosis of Graves disease was carried on 224 cases, whose treatment has been stopped for 6 months to more than 10 years. The patients were divided into 3 groups. Group 1. 92 cases, received long term antithyroid drug therapy: group 2. 100 cases, were展开更多
位姿图优化(pose graph optimization,PGO)是3D SLAM(simultaneous localization and mapping)后端优化方法之一,其精确求解依赖于良好的初始值。针对PGO噪声数据集初始化,首先提出一种新的PGO目标公式——CN(chordal with noise)模型,...位姿图优化(pose graph optimization,PGO)是3D SLAM(simultaneous localization and mapping)后端优化方法之一,其精确求解依赖于良好的初始值。针对PGO噪声数据集初始化,首先提出一种新的PGO目标公式——CN(chordal with noise)模型,此模型考虑噪声影响下产生的旋转偏差,将偏差矩阵设为参数;其次,提出ORDM(optimize rotation with the deviation matrix)算法求解CN模型,此算法在位姿图子图中,分别建立关于偏差矩阵的相对旋转测量方程,最终将CN模型化为矩阵形式,并采用线性最小二乘求出偏差矩阵的封闭解,以此修正旋转方向。实验证明,ORDM算法在面对PGO噪声数据集时,较为鲁棒,具有一定的可伸缩性;与迭代初始化算法相比,可对应较差的初始化场景。展开更多
文摘A retrospective analysis of the prognosis of Graves disease was carried on 224 cases, whose treatment has been stopped for 6 months to more than 10 years. The patients were divided into 3 groups. Group 1. 92 cases, received long term antithyroid drug therapy: group 2. 100 cases, were
文摘位姿图优化(pose graph optimization,PGO)是3D SLAM(simultaneous localization and mapping)后端优化方法之一,其精确求解依赖于良好的初始值。针对PGO噪声数据集初始化,首先提出一种新的PGO目标公式——CN(chordal with noise)模型,此模型考虑噪声影响下产生的旋转偏差,将偏差矩阵设为参数;其次,提出ORDM(optimize rotation with the deviation matrix)算法求解CN模型,此算法在位姿图子图中,分别建立关于偏差矩阵的相对旋转测量方程,最终将CN模型化为矩阵形式,并采用线性最小二乘求出偏差矩阵的封闭解,以此修正旋转方向。实验证明,ORDM算法在面对PGO噪声数据集时,较为鲁棒,具有一定的可伸缩性;与迭代初始化算法相比,可对应较差的初始化场景。