Improved 2D Color Maps and Options for Slope Map and 3D Maps


#1

When I look at trees on the elevation map, it is difficult to differentiate their height because the color becomes very dark when the height changes rapidly. Here is an example showing the tree elevation profile:

and the colored elevation-map representation of the tree:

The darkness associated with the rapid elevation change distracts from the delineation of the peak elevation of the tree. It muddies the waters, so to speak, making the sides of the tree appear higher than its peak.

Have you contemplated alternative coloring schemes that avoid this muddling? DroneDeploy produces such nice numerical data. It is a shame to that it is not best displayed in the Elevation Map.

Regards,
Terry.
@Jeremy, @Nipul, @chasemgray, @zach1


#2

Hi @SolarBarn,

Great question! To the change color scheme for areas that you zoom into, you can use the slider pictured below to give you a better representation of the elevation using the colors scheme. See below:

Please let us know if this helps.
Thank you,
@Nipul


#3

@Nipul

Yes I use this capability and like it.

My question is why does there need to be black in the picture? I understand it helps trace the topology of the various objects but it does not supply useful elevation information. The Elevation Map could just show a continuous range of colors not broken up by the grey-to-black shadings. Its like the map is schizophrenic, trying to be both a contour map and an elevation map at the same time, which has its appeal.

But I am looking for an option without the contour lines so it has continuously changing colors like this:

Regards,
Terry.


#4

Hi @SolarBarn,

I see what you’re talking about and I do see your point. This would be considered an improvement/feature.

These are great ideas and we will take them into consideration.

Thank you,
@Nipul


#5

@Nipul,

I did my own version of the 2D elevation map to show you more details.

Here is the 2D Elevation Map from DD:

And here is my 2D Elevation Map showing continuous colors without pseudo-contour lines:

Now the elevations in the trees around the buildings can be clearly seen for the range of elevations chosen. This is what I would like to see an option for in DroneDeploy.

Regards,
Terry.


#6

@Nipul,

Smooth-gradient elevation coloring also looks nice on 3D model:

Regards,
Terry.


#7

HI @SolarBarn,

Your 2D map does look very good and I’m especially liking the 3D elevation coloring. Once this feature request/improvement is prioritized, we may reach out to you for ideas.

Thank you,
@Nipul


#8

@Nipul,

I improved the coloring of the 3D elevation map to reduce the brightness variation and bring out more details:

Here is the 2D elevation map of the same area:

With the improved color-to-elevation mapping, the steps in the 2D elevation map have been eliminated and its resolution is better when the colors are centered around an elevation of interest. You can now see the mower marks in the NNW field:

And I found a way to generate the 2D elevation map with the texture of the underlying objects showing through without being too heavy handed (as in current DD implementation, IMO):

If you click on the map above to make it full screen, then it has a pseudo-3D quality in some areas due to the color choices, particularly around the tall trees in red.

Next, here is the zoomed in map from above with textures showing through:

Finally, here is my first stab at a conformal, 3D elevation map where the elevation is relative to local ground. It shows the tallest trees are down in the SE corner. Once I import LiDAR data, its accuracy will improve.

All of the maps above would look better if done inside DD, since there is higher resolution data available. My maps are reconstructed from the exported 3D model which carries only a fraction of the detail in the DD 2D maps.

Regards,
Terry.


#9

@Nipul After looking back at all of this it might also be useful in DroneDeploy if that magenta color was added part of the gamut to supply a wider spread across elevations… It could be helpful to have a little more variation so that smaller differences in elevation could be seen more easily. I have noticed this before the when trying to do a volume take off on a stockpile it is hard to tell the difference of where the actual toe of the stockpile is because the delta in the stockpile is very small in comparison to the overall site elevation delta. The only question then is that whether the magenta would be at the top or the bottom of the scale, LOL… My first thought would be the top or the highest elevation, kind of like a radar map.


#10

Yes, it is key to get as many colors as possible for maximum resolution in the elevation data. I struggle with this for my site due to its 240’ elevation change. In the top flat region, it is hard to detect the 3’ elevation change between the concrete pads of the 2 buildings. This is because the 240’ change is mapped to only 60 colors or 4’ per color! Of course this can be considerably enhanced by narrowing the viewed elevation range. In the zoomed up pictures, each color occupies only 9" and now lots of details pop.

I am trying to find more colors and will also look at your suggesting of putting magenta on top.

Do you have any pictures of what an elevation map looks like inside your Carlson tool?

Regards,
Terry.


#11

Here’s some examples of the resolution and color scheme. The first one is the entire 30ac site with 140’ of elevation change. The second is a portion of the site with 70’ of elevation change in a smaller area so the gradient is a little tighter.



#12

@chascoadmin,

These examples seem to be using only 14 colors with no blending between them so there are steps. I think this is another good option to have. I went thru this phase in my developments. Now I will go back and include it as an option. It helps to pop areas of similar elevation. Blended colors are nice where you are looking for small slopes; the color gradient makes these apparent.

The color scheme almost wraps (they totally wrap, red-to-red, on the Powerpoint Custom Color chart), with Red at one end and almost red at the magenta end. I have played with white (and grey and already use black) but I put the white above the red-yellow sequence. I should try doing it like you show here.

Looks like there is some more to play with here.

Regards,
Terry.


#13

There is a gradient option which would probably be better on a site with this much elevation change, but on our typical sites it seems to muddy the finer details. This is basically 5 feet per step and normally I might have 20 feet over the whole range so the colors would represent much smaller steps. You are correct that it does use the whole gamut of base colors. The white was an anomaly because I think it got outside of the range that I set. Your comment does bring up something though that white and black could be useful, or at least they would give me too more colors to play with. If that were the case I think I would use white at the top like you mentioned and black at the bottom. I’m going to try this site with the gradient and we’ll see what it looks like.


#14

@chascoadmin,

Ok, here’s a result with our combined color choices, white at the top and purple-red to black at the bottom:

I now have 67 colors, up 10% from 61 before. This yields 3.58’/color vs 3.93’/color before, a noticeable improvement. Plus we now have snow covered mountains!.

I think I need to tweak the colors some more in order to better equalize their perceptual differences.

Regards,
Terry.


#15

That looks amazing! I’d like to see the 2D map at about 30% capacity on the Drone deploy Ortho Mosaic as an overlay… I print mine as a PNG file out of Cad and then open it in an imaging software that can read the transparent background and erase about 60 to 70% of the layer. This allows you to put a semi transparent image and drone deploy which we have found very useful.


#16

@chascoadmin,

I am still trying to get up to speed on the various options. The use of overlays sounds interesting. Are you placing a colored elevation map out of your Carlson tool on top of the DD 2D Map? Is this what you showed in the other topic, 2D Map of Elevation Difference between Two Maps? From what it says there, it sounds like multiple overlays are involved so that you can tune the appearance. I have not yet brought up the link you included. Busy programming plus have to go to store real soon. But after I get back from the store, I will look into this.

Regards,
Terry.


#17

Yes, the images I posted are that effect. The map itself has 4 or 5 overlay layers so that you can scrutinize as seen fit. Another option I am working on with them is to be able to print (export PDF) the maps including the overlays as displayed on the screen. Here’s my process.

  1. Determine the resolution needed to balance between file size and quality
    - Templates are 2500x2500px, 5000x2500px, 5000x5000px and 10000x5000px (Rarely used, slows system)
  2. Create PNG of CAD objects
  3. Open the PNG in GIMP and verify that it has a transparent layer
  4. Select only the objects to be partially erased
  5. Adjust the Eraser Tool slide to the desired opacity and proceed erasing

#18

Hi @SolarBarn @chascoadmin,

The local recreation of the elevation map does look very good. I’m impressed. It’s great to have these examples and conversations in this thread, as this will help us when it’s time to make improvements.

Thank you,
@Nipul


#19

@Nipul, @chascoadmin, @ch

I imported the Bare Earth LiDAR data for my site and re-did the conformal 3D elevation map using our (Michael & I) combined 70 different colors. It now better clarifies the location of the tallest trees and better distinguishes their height since the 70 colors are spread across only 148’ of elevation difference or 2.1’/color.

The outside brown regions do not have LiDAR data.

For reference, here is the regular 3D elevation map with 70 colors across 240’ = 3.4’/color :

Both these maps are generated in Rhino from the exported .obj file of the DroneDeploy 3D model of this site. Rhino reads the .obj file and constructs a mesh with 500,000+ vertices which are then colored according to the elevation difference between the 3D model and the LiDAR Bare Earth base (conformal map) or just the elevation (ordinary map). Rhino can also produce a view with texture, like in the DroneDeploy 2D elevation map, but less heavy handed:

and here is the conformal version:

Regards,
Terry.


#20

Now we’re talking! Comparative datasets and then studying exactly what all can be derived from them. Is there any instance on your map where a tree or area is lower than what it previously was? Or is it all growth?