← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 8919.a · 生成模型:GPT-6

一阶二阶转向基数不同

理解等级:中级

一阶高斯方向导数由 gx、gy 两个基合成;二阶由 gxx、gxy、gyy 三个基合成。沿 t=(cosθ,sinθ) 的二阶导数为 cos²θgxx+2sinθcosθgxy+sin²θgyy。若旋转的是核坐标而非求导方向,交叉项符号可能改变,必须先定义约定。

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

      In chapter 18, we described the simplest example of an oriented filer: a Gaussian derivative. As we discussed in equation (18.27), we can synthesize a directional derivative in any direction as a linear combination of derivatives in the horizontal and vertical directions. By the linearity of convolution, that applies to the derivative applied to any filter or image, as well. It can be seen that

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第22章《滤波器组》(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/c022.xhtml) - 结论依据:摘录中关于一阶二阶转向基数不同的定义、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:In chapter 18, we described the simplest example of an oriented filer: a Gaussian derivative. As we discussed in equation (18.27), we can synthesize a directional derivative in any direction as a linear combination of derivatives in the horizontal and vertical directions. By the linearity of convolution, that applies to the derivative applied to any filter or image, as well. It can be seen that