Pointclouds in Rhino 6


I’m very excited to hear that Rhino 6 will include point cloud Import functionality. I’ve downloaded the WIP and am now in the process of running a few tests. Can someone please confirm the following for me?

Is the proposed point cloud functionality simply to be able to import various ASCII scan formats and convert the points automatically into Rhino point cloud objects? Or is a tool box/plug-in with optimised viewer, performance and editing functions proposed?

I have been using Pointools with Rhino for many years (as well as many other CAD and point cloud software tools). Despite Pointools no longer being officially supported by Rhino (I have to run the old plugin with Rhino 5 32bit Version), I still find the combination of Rhino’s modelling tools and Pointools’ point cloud performance/editing tools to provide the best solution for working with this sort of data for BIM, architectural, engineering and forensic applications. Microstation has now incorporated the Pointools engine into its software. Autodesk has followed suit by building the Recap engine into many of its software packages.

My recommendations for any potential point cloud tools in Rhino 6 would be as follows:

Useful tools found in Pointools4Rhino:

Fluid navigation around the point data with large data sets
Cut/filter data by region (box and polygon)
Hide/Show point colud
Inverse visibility of point cloud
Inverse selection of points
Deselect points
Saving regions and selection sets
Colour and intensity (shading) of points + controls of intensity Levels
Intensity in hue and greyscale
Colour by height along any axis
Snap to points
Point extraction tool (extract a number of points as a point cloud object)
Stable software – Not crashing all the time
Turn individual scans and off
Network license

Tools not in PT4Rhino that would be really useful are:

Select and filter points along a given offset to a curve – so you draw a line and say I want all points x meters either side of the line
Create mesh from point selection (using point normal if they can be imported)
Render point cloud in Rhino renderer

I hope you find this useful and would welcome the chance to offer more import or help run some tests if needed.

Please keep up the excellent work with Rhino :slight_smile:

1 Like

The first steps is to enhance display performance. See:

Next we will try to understand which tools in Pointools for Rhino are missing from core Rhino.

After that we will look at what other tools that might be needed.

1 Like

We are hoping to build whatever functionality is needed to better support point clouds into Rhino. There may be some specialized functionality that would make more sense to be written with a plug-in, but we are still at the phase of figuring out what it is that you as a user actually needs in order to work better with scan data in Rhino.

I’ll focus on the features that you listed in another reply. Thanks

1 Like

Hi All

I’ve managed to successfully import a scan file in PTS Format with colour. The pointcloud moves very smoothly within Rhino. The results are looking promising :slight_smile:

Current issues that spring forward, which I’m sure you’re already looking at:

  1. When saving the Rhino file, the size of the file is very large with the Pointcloud imported (with a single scan the file size is already 500MB). There does not seem to be a way to export the scan files in a compressed “Rhino” Format so that the scans can quickly be re-imported when needed without making the Rhino file massive. This makes data sharing quite difficult.

  2. The point data contains intensity and colour values. The colour values are displayed, but I can not see how to change the view settings to display intensity. Is this already possible?

  3. Clipping of the data is possible using clipping planes. However, this is a very limited function. Will it be possible to Clip by ploygon selection, distance from curve etc in the future?

I’ll Keep running some more tests and will report back. Keep up tehe good work :slight_smile:

Thanks and regards


1 Like

Hi Gully,

I realize this is an old thread, but would you please tell me where I can get PT4Rhino stand alone perpetual license?

Also, what software would you recommend that automates making complex surfaces out of point clouds … say, if I scan part of a boat or car sheetmetal and need to use those surfaces to make add-ons for them?

Thanx … Chris

1 Like

Hi Chris,

It’s been a while since your post so your might already found the info…

Development for PT4Rhino has been terminated since Bentley acquired Pointools.
But there is a great alternative called Arena4D for Rhino http://veesus.com/veesus/index.php/products/plugin-rhino.
It’s been over 2 years since the initial release and now it has almost everything that PT4Rhino had.

Regarding to making complex surfaces from point cloud, you might find this page useful https://wiki.mcneel.com/rhino/reverseengineering.
Relativly cheap option is RhinoResuef, if you want CAM like modelling then probably Mesh2Surface.

Now my main point, I love the massive improvement of removing points from point cloud. Now it’s much faster with Rhino 6!
My additional requests are

  • highlighting selected points like in Rhino 5 (_PointCloud_Remove)
  • unselect with Ctrl key
  • region(polygon) selection mode instead of just rectangular selection
  • LAS format support

Those would be great to see in Rhino 6!



Hi Hiro,

Thanx for getting back to me. I’m still working towards a solution to this, so your post is timely and appreciated.

I’ve looked at Arena4D, but I don’t subscribe. If they had perpetual license, I would consider getting one. Do you know of a way for me to get a perpetual license of Arena4D?

Yes, I’ve seen the reverse engineering YouTube video. I’ve read the Rhino RE article too. Done a lot of study on this subject. I know about RhinoResurf & Mesh2Surface. Thanx for the confirmations. I’ve even tried to learn Blender(!).

I ended up getting a NextEngine scanner to see how it works for my RE stuff. Turns-out it’s not all it’s cracked up to be. I wasted 2 long weeks trying to get the scans to align properly w/their ScanSoft software, and then with their RapidWorks software, but nothing doing. A veteran RE guy I talked to (from aerospace industry) told me it’s caused by deficiencies in the software.

So basically, what I need now, is a point cloud processor to assemble multiple scans accurately. Then I can use the point cloud in Rhino for post processing. Then I can take the construction geometry into 3D CAD to do what I want.

One thing’s for sure: There’s a lotta extra work involved in doing RE work.

Thanx for the input.
Any further comments or suggestions are welcome.

Cheers … Chris

Hi Chris,

Arena4D started as perpetual but became subscription last year… I’m not sure whether you can still get perpetual license anywhere, I guess the best shot would be taking to resellers in your country or Veesus directly.

Either way, Arena4D doesn’t do point cloud registration it won’t be useful for your purpose. In fact, I don’t think there are any point cloud registration plug-ins for Rhino…

Have you tried freeware called CloudCompare? <http://www.danielgm.net/cc/> It doesn’t register multiple clouds at once, but you can align 2 clouds. So if you don’t have too many point cloud bits, you can align one by one then merge in the end. I tired with TLS point clouds and the processing time each was pretty quick. You might need to roughly align clouds manually first before the Fine registration(ICP). Make sure you have sufficient overlaps (maybe 20%?) in the clouds that you are trying to register.
The exported point cloud can be imported into Rhino for modelling process.

Give a go if you haven’t tried CloudCompare. Otherwise, my next suggestion is Geomagic which is very pricey…


1 Like

Hi Hiro,

Thanx for the info.

I tried Blender and Cloud Compare. The learning curve for Blender is too steep. Clound Compare isn’t robust enough. Also, Blender’s interface is hardly ‘intuitive’. It’s nothing like the norm in the computer world. Why someone would write software so different from the norm is crazy. If they want people to use it, they need to make it the same as the norm. Anyways …

I usually have 30 to 50 scans that need aligning. In the future, it could be 100’s. I’ve used Design-X to align sheetmetal scans. It failed! It’s GUI and work-flow is very old compared to the CAD I presently use (too many keystrokes). It certainly isn’t worth $20k! Maybe more like $4k. The only thing that looks inviting to me is Creaform, which uses targets. According to a guy I talked to in the aerospace industry, Creaform uses one of the most accurate methods to create a scan and align it properly. He said it’s good to .0001"/.0025mm! But it’s WAY too expensive for my purse!

Never heard of TLS. I’ll look into it. Thanx.

Ever hear of David? I tried to find the free version, but since it HP bought it, they apparently cleaned all free versions from the internet. Too bad, b/c it looked very promising. I was one year too late in getting it. Otherwise, I’d have tried David. Do you know where I can get a free version of it before HP bought it?

What scanner do you use to do laser scanning? What kinds of scanning do you do? What software do you use?

Cheers … Chris

Hi Hiro - that one is on the bug pile. SelBrush SelBoundary, Lasso and other seleection tools should work now in Remove, but the highlight problem makes it difficult to use…

I also see that a pre-selection is ignored with Add/Remove, which seems wrong to me.



Hi Chris,

TLS (Terrestrial Laser Scanner) is for scanning building and landscape for surveying purposes.
To be honest I works with TLS and Mobile Laser (scanners on a car) mainly, and only scratched a surface of close-range scanning & RE modelling (Faro Arm & Rapidform before merged to Geomagic). …Sorry I wasn’t helpful after all :frowning:

I hope you will find ways without spending too much $$$!


Hi Pascal,

Thanks for the tips and noting the bug. Lasso is exactly what I wanted :wink:
I’ve created a new post for this, and hoping someone can comment on LAS support there…


Hi Hiro,

What do you do with the mobile scanner?

Cheers … Chris

This is my first post here but have been using Rhino for years. I created an account just to second everything Gully10 has requested concerning point clouds. A Pointools-like functionality within Rhino would be a huge relief.

Point clouds are becoming more and more common in work flows and Rhino should not be left in the dust here by its competitors. Rhino is the best tool for reverse engineering most things because of its flexibility and point clouds are becoming a necessity.

We need more import options, lighter files, color and intensity sliders, slices, slices from curves, etc etc etc. We basically need Pointools or Arena 4d to be native in Rhino. You’d have a product that was much more marketable to the reverse engineering, laser scanning, architecture, BIM, etc. communities. That and I could retire my Pointools dongle before it dies on me!

Thanks, and fill me in if any of this has been added to Rhino 6 and I’m missing it.


I really don’t want to get creative and make up work to do😄 Do you have a specific request for working with point clouds in Rhino that needs to be addressed so you can get your work done? “Make it work like pointtools” is too broad of a request, so narrowing in on something specific really helps us understand what needs to be done.

Thanks for responding Steve. Sorry if I was brief but I just wanted to chime in so you guys know that there are more of us out here who would like to see these changes.

If you were to make my dreams come true, ideally, I’d like a dedicated panel or tab for point clouds which has a myriad of useful point cloud related tools. I’ll list my biggest wishes.

  • Visual sliders that I can grab and pull that show only a slice of the point cloud at a predetermined thickness. This would be a c-plane and view oriented to that c-plane that I can quickly and easily manipulate (or turn on and off if need be). Being able to slide that slice along a path or curve would also be highly useful. This way I can work fast and easily, and spend more time modeling or drawing, rather than orienting my view. If I have a point cloud of a building and need to do a plan view drawing of each floor, it would be nice to just grab a slider for the top view in, say, the front view port and drag it down to about a foot off of the ground on whatever floor I need to draw.

  • The ability to turn color/intensity on and off and even play with their levels. Also the ability to add color to point clouds that have no color or intensity. In Pointools, for example, you can add a hue spectrum to the point cloud and adjust the axis that spectrum displays on, whether the spectrum repeats or stretches over the whole cloud, and the overall scale of the spectrum. This makes it much easier to see and work with clouds that do not have intensity or color.

  • A simple set of point cloud dedicated selection tools. This is convenient for focusing on specific parts of point clouds without having to hide, or accidentally grabbing your work. A polygonal selection tool would be ideal, and from this set of tools you could select, deselect, hide, or reveal points.

  • A lightweight native Rhino point cloud format. Pointools does this with .pods, and Autodesk does it with .rcp (recap). This may or may not be necessary but the issue is that massive point clouds just don’t handle well. Some way to make very large point clouds handle easier and import faster would be great. Forgive me because I don’t know too much about the details on this one.

I understand that some of these functions are already in Rhino, but having a more robust point cloud section would be wonderful. Something dedicated to importing, manipulating, and modeling/drawing from point cloud data of all sizes and types.

Thanks for your time.

I agree with all the above comments.

The big thing for me is being able to select only the points I need in a huge point cloud and the process them.

The next biggest thing is to be able to process those points using the primitive 3d shapes, such as PlaneThroughPt, but also have available every other primitive shape for use with the point selection.

Then I need to be able to orient the model to the World Coordinate system in order to create orthogonal features out of the points.

I just got through testing Mesh2Surface and found it excellent, but still lacking in a number of ares. But it doesn’t handle point clouds. So I’m sort’a stuck in limbo trying to get point clouds into workable features for 3d parametric CAD.

One example of a need is to create a work plane mid-way between 2 non-planar surfaces that are mirrors of each other. The key in working with these point clouds (or meshes) is being able to create construction features to import into 3d parametric cad, b/c that’s where I create the 3d model and then the fabrication drawings (esp. when doing sheetmetal parts).

Also, Rhino crashes with an “out of memory” error every time I try to explode a huge point cloud, in order to get access to select only the point groups needed for processing. It reminds me of Dos 2.X and Windows 3.X days with AutoCAD!

Cheers …

Would ‘points to cad’ be too much to ask ;)))
See this video at ca. 0:55

1 Like

Yeah, except the accuracy is 3/8" in 3 feet, which is WAY too high a tolerance for building. When I made cabinets, 1/8" in 6’ was acceptable, but not across a whole room, only on one spot were a blemish was. So the advertising and marketing is a lie. In indusrial CAD, 1/64" in 6’ is barely acceptable!

Still, “points to CAD” would be nice.

I just think criminals at the corporations in ‘the industry’ are milking the market for all they can get out of it. That’s how it works today. You create a market, then you manage that market, then you dump you product when the charts show you that your market is saturated. I believe the term for that is EXPLOITATION (of society). Another term for it is SOCIOPATH (due to the damage done to the social institutions and environs). Just cuz it’s common custom doesn’t make it right.

1 Like

Hi Pascal,

Point clouds with normal disappear when they get selected in Shading & Rendering display modes, instead of being highlighted. This was persistent in at least last 2 builds, and the current one I’ve just downloaded.
Do you have any tips to solve this problem?