与SURF匹配（概念）
我正在使用 surf 将 n 个图像匹配在一起。通过在匹配阶段去除异常值来找到基本矩阵。令人惊讶的是，即使匹配点完全错误，我也找到了基本矩阵。我的问题是我应该使用什么标准或静态来避免此类问题。 问候，
I am using surf to match n images together. finding fundamental matric by to remove outliers in matching stage. Surprisingly i am finding fundamental matrix even though matching points are totally wrong. My question is what criteria or statics i should use to avoid such problem.
regards,
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（1）
您的代码示例或您所需内容的详细说明可能会帮助我们向您展示一条路径。
如果你的意思是，即使在匹配两个相同的图像之后，匹配的点也是错误的，那么它们在描述符的计算中就有可能出错。至少您得到的错误的示例输出会对我们有很大帮助。关键点是否被正确检测？或者只是匹配出错了？
An example of your code, or a detailed explanation of what you require might help us to show you a path.
If what you meant is that, even after matching two same images, the points being matched are wrong, then their is a chance for error in the calculation of descriptors. At least a sample output of the error you get would help us a lot. Are the keypoints are being detected correctly? Or is it just that the matching is going wrong?