← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 8885 · 生成模型:GPT-6

逐点乘法在频域变成卷积

理解等级:高级

空间中两个信号逐点相乘,对应频谱圆周卷积,并带本章归一化系数一维1/N、二维1/(NM)。因此用固定空间遮罩截取图像,也可能扩散或混合频率。固定遮罩对输入仍是线性的,频率混合不必一律来自非线性。

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

      It turns out that the Fourier transform of the product of two images is the (circular) convolution of their DFTs: (16.47) We leave the proof of this property to the reader. 16.7.6 Shift Property, Translation in Space A shift corresponds to a translation in space. This is a very common operation that has a very particular influence on the Fourier transform of the signal. It

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第16章《傅里叶分析》(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/c016.xhtml) - 结论依据:摘录中关于逐点乘法在频域变成卷积的定义、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:It turns out that the Fourier transform of the product of two images is the (circular) convolution of their DFTs: (16.47) We leave the proof of this property to the reader. 16.7.6 Shift Property, Translation in Space A shift corresponds to a translation in space. This is a very common operation that has a very particular influence on the Fourier transform of the signal. It