知识卡片
盒核先是求和归一化后才平均
单位矩形核将邻域像素求和,直流增益等于核内样本数。半径N、M的二维盒核应除以(2N+1)(2M+1),才成为保常量的均值核。核尺寸、归一化与边界扩展都需记录,不能把未归一化求和直接称亮度不变模糊。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 模糊滤波器
原书出处 · 1 处
- · 查看原书依据
That is, the output value on each location (n, m) is the sum of the input pixels within the rectangle around that location. Visually, filtering an image with the box filter results in blurring the picture. Figure 17.2 shows some box filters and the corresponding output images (after normalizing the box filter coefficients so that they sum to 1). Figure 17.2(a) shows an image convolved
参考来源
- 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第17章《模糊滤波器》(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/c017b.xhtml)
- 结论依据:摘录中关于盒核先是求和归一化后才平均的定义、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:That is, the output value on each location (n, m) is the sum of the input pixels within the rectangle around that location. Visually, filtering an image with the box filter results in blurring the picture. Figure 17.2 shows some box filters and the corresponding output images (after normalizing the box filter coefficients so that they sum to 1). Figure 17.2(a) shows an image convolved