知识卡片

普通读书笔记卡 · 8903.b · 生成模型:GPT-6

二维混叠会改变方向

理解等级:中级

高频二维条纹被过稀网格采样后,不仅看起来频率降低,主导方向也可能改变。这样的结构并非原图中真实存在的粗纹理,而是不同频谱副本叠加后的结果。图像识别前的缩放会因此改变纹理证据,不能把分辨率下降只理解为细节变少。

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

      Let's now look at an example in two dimensions (2D). As a concrete example, let's consider the 2D signal shown in figure 20.4(a). This 2D signal is a continuous function in the spatial domain defined as ℓ(x, y) = cos(w(x + 2y)) and with a frequency, w, which also varies with location following w = x + 2y. That is ℓ(x, y) = cos (x

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第20章《图像采样与混叠》(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/c020b.xhtml) - 结论依据:摘录中关于二维混叠会改变方向的定义、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:Let's now look at an example in two dimensions (2D). As a concrete example, let's consider the 2D signal shown in figure 20.4(a). This 2D signal is a continuous function in the spatial domain defined as ℓ(x, y) = cos(w(x + 2y)) and with a frequency, w, which also varies with location following w = x + 2y. That is ℓ(x, y) = cos (x