知识卡片
兴趣点需要跨视图重复性
兴趣点既要局部可定位,还要在尺度、视角和照明变化后能被重复找到。强梯度只是部分条件,反光、重复图案或只出现一图的细节仍可能不能可靠匹配。评估检测器不能只数点多,还应检查轨迹重复和几何一致性。
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
原书出处 · 1 处
- · 查看原书依据
The goal of finding image features is to locate image locations that are likely to be stable under different sets of geometric transformations (translation, scaling, rotation, skew, and projective), and illumination changes. These stable regions are called interest points.
参考来源
- 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第44章《多视图几何与运动恢复结构》(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/c044.xhtml)
- 结论依据:摘录中关于兴趣点需要跨视图重复性的定义、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 goal of finding image features is to locate image locations that are likely to be stable under different sets of geometric transformations (translation, scaling, rotation, skew, and projective), and illumination changes. These stable regions are called interest points.