Hello,
I am currently looking for a way to find out, whether 2 polylines are touching, or crossing each other. I have tried the "vla-intersectwith" method, but it returns an output in both cases. I don't want to use bounding boxes, as it would give false results, depending on the polylines' shapes.
I've spent a few hours on google, looking for an answer, but didn't find anything. I would be very thankfull if you could send some ideas my way, on how to approach the matter.
Thank you.
That idea went through my mind at one point, but i wrote it off, because it will only work for rectangles. A whole parameter value indicates a vertex, which would work in cases, where one of the polylines touches the other with a vertex /or multiple verteces/.
Maybe my initial image wasn't a good example. I want to find out, whether the 2 plines are touching each other, regardless if it's at a vertex or not.
First of all i want to thank you for your ideas. I really appreciate your help.
Maybe my brain isn't working at the moment, so i can't quite get your idea. (- par (fix par)) would give me the parameter of the touch/cross, for the pline segment. At the moment i can't think of a way that is usefull for me.
Btw i just made another example of touching and crossing plines, both with 2 points, and both not on verteces, and i can't think of a way do differentiate the two.
Are the likely shapes "ordinary" enough, and is the degree of overlap [if any] smallish compared to the sizes, as in your examples, and would there not be other stuff around inside the shapes? If so, you could find the centroids or middles of bounding boxes, use those as internal-pick points for BOUNDARY commands, and compare the areas of the results to the areas of the originals. If the originals touch but don't overlap, there will be something returned by the (intersectwith) method, but the BOUNDARY results will have the same areas as the originals, no matter how many points of intersection or how much length of coincident edge(s) there may be. If the originals overlap, the BOUNDARY results would each have a smaller area than its corresponding original, no matter how many points of intersection there are.
First, check for intersections with vla method.
Now check if the points between every 2 consecutive intersections of one polyline are inside/outside the other polyline.
Based on the alternation inside/outside, you can establish the relative position of the 2 polys.
Now you only need a good inside_p function 🙂 I guess you'll find some in this forum.
See the image for demonstration. Notice it doesn't matter which polyline is checked against the other.
Further questions....
Are you talking about only and always closed Polylines? Or other closed shapes [Circles, Ellipses, Splines]? If so, there's one definitive condition: if there's only one intersection, that means they touch but do not overlap, so no further analysis would be needed.
And what about shapes that overlap but do not intersect?
@doaiena wrote:
....
In the 1-st and 3-rd images in this thread, are cases where the plines produce 2 intersection points, both for crossing and touching.
Yes, and the upper one in Message 3 is what I was talking about -- one intersection, so you don't need to bother analyzing whether they simply touch or overlap [as you would if there are more than one intersection], because if they're both closed, one intersection can only mean that they touch without overlapping.
