知识卡片
RANSAC内点阈值决定共识
阈值太小会把真实噪声点排除,太大则可能把其他结构误当支持。共识最多不一定是任务要找的对象,还要看模型、空间范围和数据解释;RANSAC给概率性搜索手段,不保证任意高离群比例下都找到真解。
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
- · 特征提取
原书出处 · 1 处
- · 查看原书依据
A powerful solution to this problem is to randomly sample possible lines and keep those that satisfy a certain desired quality, given by the number of points being somewhat close to the best fit. This is illustrated in figure 9.3, with darker lines corresponding to better fits. RANSAC usually requires two parameters—namely, the number of points required to consider a line to be a valid
参考来源
- 位置:《Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)》第09章《特征提取》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)/_epub-src/OEBPS/xhtml/chapter_9.xhtml)
- 结论依据:摘录中关于RANSAC内点阈值决定共识的定义、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:A powerful solution to this problem is to randomly sample possible lines and keep those that satisfy a certain desired quality, given by the number of points being somewhat close to the best fit. This is illustrated in figure 9.3, with darker lines corresponding to better fits. RANSAC usually requires two parameters—namely, the number of points required to consider a line to be a valid