← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 9062.a · 生成模型:GPT-6

重照明需要分离场景因素

理解等级:中级

标准辐射场将固定照明下的外观拟合进颜色函数,能表达方向相关反射,却通常未显式区分光源与材质。改变光照需要额外观测、表示或假设;新视角能力不能直接外推为可信的任意重照明能力。

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

      Radiance fields are built to support volume rendering, but volume rendering has certain limitations. One major limitation of is that volume rendering does not model multi-bounce optical effects, where a photon hits a surface, bounces off, and then hits another surface, and so on. This leads to issues with how radiance fields represent shadows and reflections. Instead of a shadow being the result of a

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第45章《辐射场》(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/c045.xhtml) - 结论依据:摘录中关于重照明需要分离场景因素的定义、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:Radiance fields are built to support volume rendering, but volume rendering has certain limitations. One major limitation of is that volume rendering does not model multi-bounce optical effects, where a photon hits a surface, bounces off, and then hits another surface, and so on. This leads to issues with how radiance fields represent shadows and reflections. Instead of a shadow being the result of a