← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 8895.d · 生成模型:GPT-6

二项式导数的离散替代

理解等级:中级

采样连续高斯导数是一种离散化方案;另一种是在二项式平滑核上再卷积 [1,−1]。后者给出短小、整数系数的差分平滑组合,便于高效计算。不同核的支撑中心与归一化不同,尤其比较奇偶长度核时应把空间偏移一并记录。

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

      When processing images we have to use discrete approximations for the Gaussian derivatives. After discretization, many of the properties of the continuous Gaussian will not hold exactly. There are many discrete approximations. For instance, we can take samples of the continuous functions. In practice it is common to use the discrete approximation given by the binomial filters. Figure 18.13 shows the result of convolving the

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第18章《图像导数》(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/c018.xhtml) - 结论依据:摘录中关于二项式导数的离散替代的定义、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:When processing images we have to use discrete approximations for the Gaussian derivatives. After discretization, many of the properties of the continuous Gaussian will not hold exactly. There are many discrete approximations. For instance, we can take samples of the continuous functions. In practice it is common to use the discrete approximation given by the binomial filters. Figure 18.13 shows the result of convolving the