知识卡片
前向投点容易留空洞
前向映射遍历源像素,将变换后的坐标舍入后写入目标网格。放大、旋转等情况下可能多个源点落同格,也可能某些目标格无人写入,产生冲突和空洞。此处问题来自简单投点栅格化;更复杂的面积投影或splatting需额外规则,不应和简单取整混为一谈。
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
- · 图像与几何表示
原书出处 · 1 处
- · 查看原书依据
The first algorithm that usually comes to mind when transforming an image is to take every pixel in the original image represented as [ℓi, xi, yi], apply the transformation, M, to the coordinates, and record the pixel color into the resulting coordinates in the target image (figure 38.10). As coordinates might result in non-integer values, we can simply round the result to the closest pixel
参考来源
- 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第38章《图像与几何表示》(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/c038b.xhtml)
- 结论依据:摘录中关于前向投点容易留空洞的定义、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:The first algorithm that usually comes to mind when transforming an image is to take every pixel in the original image represented as [ℓi, xi, yi], apply the transformation, M, to the coordinates, and record the pixel color into the resulting coordinates in the target image (figure 38.10). As coordinates might result in non-integer values, we can simply round the result to the closest pixel