← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 8902.d · 生成模型:GPT-6

多速度局部零化有边界

理解等级:中级

零化滤波器可用于具有多个运动目标的视频,只要其局部支撑内近似只有一种速度,匹配成分就会被抑制。遮挡边缘、多运动混合、速度变化和核截断会破坏理想零化。透明图层若按亮度线性相加可用叠加原理分析,不能直接套用于乘性透明或复杂遮挡合成。

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

      Note that despite that we assumed that the sequence contained a single moving object when deriving the formulation for the nulling filter, the filter also works in sequences with multiple objects moving a different speeds as shown in figure 19.7. This is because the operations are local (the kernel h has a small size) and the behavior will be correct if in a local image

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第19章《时间滤波器》(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/c019.xhtml) - 结论依据:摘录中关于多速度局部零化有边界的定义、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:Note that despite that we assumed that the sequence contained a single moving object when deriving the formulation for the nulling filter, the filter also works in sequences with multiple objects moving a different speeds as shown in figure 19.7. This is because the operations are local (the kernel h has a small size) and the behavior will be correct if in a local image