知识卡片

普通读书笔记卡 · 9025.c · 生成模型:GPT-6

上三角正交分解恢复内外参

理解等级:高级

B=KR 需上三角 K 与正交 R,可对B作RQ分解,或若 B^−1=QU,则 K=U^−1、R=Q^T。还需处理M的尺度、K归一化与轴约定,使R属于合法旋转。只分解数值矩阵不保证分解解释符合实际相机。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      The next step is to decompose B into the product of two matrices KR. These two matrices are very special. The matrix K is an upper triangular matrix by construction (in the case of perspective projection), and R is an orthonormal matrix. To estimate K and R we can use the QR decomposition of the matrix B–1. The QR decomposition decomposes a matrix in the

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第39章《相机建模与标定》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c039.xhtml) - 结论依据:摘录中关于上三角正交分解恢复内外参的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。 - 原始内容:The next step is to decompose B into the product of two matrices KR. These two matrices are very special. The matrix K is an upper triangular matrix by construction (in the case of perspective projection), and R is an orthonormal matrix. To estimate K and R we can use the QR decomposition of the matrix B–1. The QR decomposition decomposes a matrix in the