A summary of methods for finding the boundary of point clouds

In real life, most of the time we collect data is finite point data, interpolation can generate a polygon layer. The general step is to use the boundary of the point cloud to trim the interpolated result data. Here is a summary of the method of finding the minimum boundary of a point cloud.

Method One: Manual vector method

If the amount of data is small, manual vectorization, accurate and simple.

cons : With a large amount of data, it's a bit annoying.

Method Two: Minimum geometric boundary

This is the easiest and simplest way to operate, and this tool is directly available in ArcGIS. If the requirement is not very high, this method can achieve the goal. Specific methods refer to the official website

Disadvantage : You cannot create concave polygons unless you generate a convex polygon.

Method Three: Tin Transfer method

Everyone can try this method, I have not tried.

http://blog.csdn.net/esrichinacd/article/details/7788065

Method Three: Point density method

Http://bbs.3s001.com/thread-148639-1-1.html

I do not recommend using this method.

Method Four: Aggregation Point method

Here is an analysis of this method. The function of the aggregate points tool in ArcGIS is to create polygon features around approximate point feature clusters. It takes at least three points to become polygons, so it is a polygon created around a cluster of three or more points within the aggregation distance range.

So, the key to using this approach is that how to set the minimum aggregation distance is critical. Let's take a look at the effects of different aggregation distances.

150 and 250 units of the processing results, we can see that two graphs have missed a point, and the left side of this more contour. In order not to lose the data, we have to increase the distance.

The results of the processing of 280 and 500 units respectively, we can see that two graphs are not leaking points, and the left side of this more contour. In order not to lose the data and the boundary is more contoured, we have to find a threshold.

To find the critical value, it is important to understand the words in the Help document- creating polygons around clusters of three or more points within the aggregation distance , that is, the previous point is not included because there are no three points in that distance, as shown, to be included in the 9th point, We need to set the aggregation distance should contain 184, 185, 203 points, that is, 9th to the minimum distance of three points of the maximum value, a bit awkward. By far, that is to find the shortest distance of all points, the last three points, and then select one of the maximum values, which is the threshold we need.

This can be achieved using the Generate near Table tool in ArcGIS. After generating the minimum boundary, I also made a buffer and smoothing. Next in is generated:

A summary of methods for finding the boundary of point clouds