用于从轮廓进行表面重建的库或算法 [英] Library or algorithm for surface reconstruction from contours

问题描述

大家好，

从轮廓（闭合多边形）中寻找表面重建的算法或库。

我想创建围绕两个多边形的最佳曲面。每个多边形都有一些凸凹部分;因此，从多边形重建表面更加困难。我正在寻找一个算法或库或任何想法来解决这个问题。



数据的例子（两个多边形）如下：

https://www.dropbox.com/s/v9ccblqw96jgvf9/Data-Example。 jpg？dl = 0 [ ^ ]



表面的一部分，我想重建如下图所示：

< a href =https://www.dropbox.com/s/ed02t6xeg4wmq1y/Data-Example2.jpg?dl=0> https://www.dropbox.com/s/ed02t6xeg4wmq1y/Data-Example2.jpg?dl = 0 [ ^ ]



对此的任何意见表示赞赏

谢谢



我尝试了什么：

根据文献，有一些方法例如通过最小化成本函数（例如，最小化总表面）的优化方法，但是我发现的方法对于我的数据不够好，哪些多边形具有多个凸面和凹面部分。发现的方法可以追溯到20  -  30年前。我想知道，是否有一种新开发的方法可以更好地处理这个问题。我尝试最小化总面积方法，但这种方法对我的数据不起作用。

解决方案

https://www.dropbox.com/s/ v9ccblqw96jgvf9 / Data-Example.jpg？dl = 0 [ ^ ]

据我所见，点密度太低。

如果你提供一个多边形的点对于期望单个多边形（如TSP）的算法，可能会重建凹形轮廓。

如果同时提供多边形点或算法不期望单个多边形，则重建将失败

Hello everyone,
Looking for an algorithm or library for surface reconstruction from contours (closed polygons).
I want to create the best surface surrounding two polygons. Each polygon has some convex and concave sections; therefore, reconstruction the surface from polygons is more difficult. I am looking for an algorithm or library or any idea to solve this problem.

An example of data (two polygons) is as below:
https://www.dropbox.com/s/v9ccblqw96jgvf9/Data-Example.jpg?dl=0[^]

A part of the surface, I would like to reconstruction is available in the following figure:
https://www.dropbox.com/s/ed02t6xeg4wmq1y/Data-Example2.jpg?dl=0[^]

Any input on this is appreciated
Thanks

What I have tried:

based on the literature, there are some methods such as optimization methods by minimizing a cost function (for example, minimizing the total surface), but the methods I found does not work for my data good enough which polygons have several convex and concave sections. The methods found comes back to 20-30 years ago. I am wondering, is there a newly developed method to handle this problem better.I tried minimizing total area method, but this method does not work well for my data.

解决方案

https://www.dropbox.com/s/v9ccblqw96jgvf9/Data-Example.jpg?dl=0[^]
As far as I can see, points density is too low.
If you provide points of a single polygon to an algorithm that expect a single polygon (like TSP), the concave contour will probably be reconstructed.
If you provide both polygons points together or if algorithm do not expect a single polygon, the reconstruction will fail.

这篇关于用于从轮廓进行表面重建的库或算法的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！