← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9073.b · 生成模型:GPT-6

反向光流不是原地取负

理解等级:中级

前向流在p上定义,将其映到q=p+f(p);有效可逆对应处的反向流应满足b(q)=−f(p)。它定义在第二帧坐标,不等于一般地对同一数组索引的前向流取负;遮挡和多对一映射会进一步破坏简单逆关系。

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

      Optical flow is an approximation to the 2D motion field computed by measuring displacement of image brightness (figure 48.1). The ideal optical flow is defined as follows: given two images ℓ1 and ℓ2 ∈ ℝN×M×3, the optical flow [u, v] ∈ ℝN×M×2 indicates the relative position of each pixel in ℓ1 and the corresponding pixel in ℓ2. Note that optical flow will change if we

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第48章《光流估计》(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/c048.xhtml) - 结论依据:摘录中关于反向光流不是原地取负的定义、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:Optical flow is an approximation to the 2D motion field computed by measuring displacement of image brightness (figure 48.1). The ideal optical flow is defined as follows: given two images ℓ1 and ℓ2 ∈ ℝN×M×3, the optical flow [u, v] ∈ ℝN×M×2 indicates the relative position of each pixel in ℓ1 and the corresponding pixel in ℓ2. Note that optical flow will change if we