知识卡片

普通读书笔记卡 · 8839.f · 生成模型:GPT-6

正交模型是有条件的透视近似

理解等级:中级

正交投影可写x=kX、y=kY,缩放k固定、不随深度改变。若物体相对深度变化远小于整体距离,可把透视的f/Z近似为常量。远摄本身不保证所有远近物体同样大小;有限距离的相机仍有透视,只在所考察深度范围内近似。

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

      where the constant scaling factor k accounts for change of units and is a fixed global image scaling. Orthographic projection is a good model for telephoto lenses, where the apparent size of objects in the image is roughly independent of their distance to the camera. We used this projection when building the simple visual system in chapter 2. Orthographic Projection is a correct model when

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第05章《成像》(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/c005b.xhtml) - 结论依据:摘录中关于正交模型是有条件的透视近似的定义、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:where the constant scaling factor k accounts for change of units and is a fixed global image scaling. Orthographic projection is a good model for telephoto lenses, where the apparent size of objects in the image is roughly independent of their distance to the camera. We used this projection when building the simple visual system in chapter 2. Orthographic Projection is a correct model when