OPENCV:使用FindCirclesGrid检测大型校准模式
我正在使用OPENCV检测校准目标(特别是不对称圆圈,但该问题也适用于对称圆圈)。
当特定圆圈中的像素数很小(从经验中,直径小于大约30-40像素)时,FindCirclesGrid的工作正常,但是如果直径太大,则不会返回。我已经尝试创建不同尺寸的目标(感谢, calib.io ! ),整体网格大小似乎无关紧要,但是如果每个圆圈太大,则检测会失败。
我假设对Contours / Blob型函数进行了一些调用,正是由于某些最大尺寸约束而导致的呼叫。是否可以在过程的轮廓/斑点部分中指定参数,而无需重写整个FindCirclesGrid?我在Python工作,因此任何基本的重写都在编程时间和运行时效率低下。
任何帮助将不胜感激。
I'm using OpenCV to detect calibration targets (specifically asymmetric circles, but the question applies to symmetric circles too).
When the number of pixels in a particular circle is small (diameter less than about 30-40 pixels, from experience), findCirclesGrid works fine, but if the diameter is too large None is returned. I've tried creating targets at different sizes (thanks, calib.io!), and the overall grid size seems irrelevant, but if each circle is too big, the detection fails.
I assume some calls are made to contours / blob-type functions, and it is these which are failing due to some max size constraint. Is there a way to specify parameters in the contour/blob part of the process without rewriting the entire findCirclesGrid? I'm working in Python, so any fundamental re-write would be inefficient in programming time and run-time.
Any help would be appreciated.
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