Hi,
I am wondering how to perform 3D viewshed analysis in the built environment (as the attached picture). Is there any plug-in able to do this?
Best,
-HC
The resolution slider added in the next post makes this a versatile tool. My laptop now is three times faster running this model than what I had then (July 7, 2016).
flashlight_2016_Jul7b.gh (64.3 KB)
2 Likes
This is an updated version of the old model that now uses a spherical projection of points with sliders for ‘Radius’, ‘Resolution’ (smaller value is more points) and ‘Tolerance’ (increases cull radius). Rays that miss all obstacles and reach the sphere are culled.
flashlight_2020Oct7a.gh (56.3 KB)
1 Like
Added a Data Dam (red group) that blocks green group and slow IVRay
while adjusting ‘Choose Viewpoint’ and ‘Two Hill Surface’. (updated)
flashlight_2020Oct7c.gh (62.9 KB)
1 Like
Amazing that you provide lots of information for the view analysis. I am fairly new to grasshopper and it takes time to understand your code. But I got some questions to ask with:
-
When I try not to create a terrain in the GH, I disenable the definition. But then it doesn’t work. I am just wondering what if I already have buildings and terrain for the analysis, which part of the definition I should keep for the analysis? (attached figure 1)
-
I tried to draw some lines to represent the visibility of the viewpoint. (attached figure 2) But I have no idea to do with those BLUE points. Is it possible to draw lines to represent visibility?
Best,
-HC
Figure 1:
Figure 2:
The Data Dam (red group) in flashlight_2020Oct7c.gh are the three inputs you want.
- ‘A’ is a list of obstacles.
- ‘B’ is the viewpoint.
- ‘C’ is a surface (that gets combined with the list of obstacles).
1 Like
Instead of Data Dam I have split the code into two pieces connected by R6 Data Output and Data Input, one that generates the geometry (View_Pt, Obstacles and srf) and the other for the IVRay treatment with some cosmetic visuals.
ivray_geo_2020Oct8a.gh (50.3 KB)
ivray_2020Oct8a.gh (25.6 KB)
1 Like
Manual adjustment of ‘Two Hill Surface’ group…
Resolution = 10, Tolerance = 10, point count = 114150, CullPt = 50 secs, IVRay = 1.6 mins.
Much quicker at Resolution = 20, Tolerance = 20.
1 Like
DecodingSpaces can do it with GPU accleration, but it frequently crashes when there’s a ton of data.
Mine is slightly faster than Joseph’s but less accurate.
2 Likes
Thanks Joseph,
Trying my best to understand the definition!
-HC
Hi @gankeyu,
May I know what are these components? Is it a plugin?
I did try DecodingSpace, but as you say, It is easy to crash.
Best,
-HC
Would like to see your list of viewpoints, obstacles and optional surface instead of my own test geometry from ivray_geo_2020Oct8a.gh ?
ivray_2020Oct12a.gh (28.5 KB)
ivray_geo_2020Oct8a.gh (50.3 KB)
1 Like
Hi @Joseph_Oster,
I am trying out the different plugin and your definition. Here is the conclusion so far (I am still working on it):
-
Ladybug: it is easy for view analysis but the outcome is like the first figure I share. (And also it should have both obstacles and geometry for input. The result will not be accurated if I only test viewpoint and terrain geometry)
-
Decoding space: similar to ladybug but more complicated for new users.
-
Your definition: Still trying to figure out the code.
Attached is the geometry and viewpoint I am trying to test.
Best,
-HCterrain model.3dm (5.8 MB)
Your geometry, my code:
ivray_geo_2020Oct13a.gh (1.8 MB)
ivray_2020Oct13a.gh (27.2 KB)
Boxes are arbitrary size, making “obstacles” of other view points.
1 Like
Hi @Joseph_Oster,
Thanks for your answer. It inspired me for another kind of view analysis (attached figures) and hence I have some more idea about the view analysis:
-
I am just wondering if I don’t need to quantify how much the target that can be seen, but only consider which viewpoint can see the target and which not. The result can just illustrate by a simple line (blue line), which means it can indicate which part of the terrain blocks the view.
-
Another inspiration is the landmark of Chicago, which can reflect the city view by a bean. And I think this kind of view analysis could be the best representation to see what view can be concentrated in the bean. (attached figure)
Let’s say the bounding box represents the bean, and I would like to see what view can be concentrated in the box. Do you have any idea about this?
1014_terrain model.3dm (5.7 MB)
Best regards,
-HC
Hi @Joseph_Oster,
It is a bit complicated for me. But I write a simpler definition for tracing the line. The problem is I can’t remove the line that didn’t touch the building surface (I just want the line which touches the geometry).
1015_3D_isovist_HC.gh (6.8 KB) test geometry.3dm (44.3 KB)
Best,
-HC