Looking to create 3D model of building.


 Share

Recommended Posts

Ok this may me a dumb question but here goes. I use PTGUI to create interactive 360x180 Pano's. Has anyone tried using PTGUI to do the opposite? fly around a building instead of drone static taking 360 photos. I'm looking for stand alone software to do 3D modeling. Don't want to do the web based thing.

Link to comment
Share on other sites

  • 1 month later...

I've tried using PTGUI and Stitcher.  Neither work well for Ortho-rectified mosaic.

 

You might try ContextCapture [used to be Acute3D] from Bentley Systems.  Pix4D is OK if your not looking for accuracy and most of the online systems like DroneDeploy and DataMapper don't give you enough control over ground control points to get accurate maps.  They look good though...

I'm looking for software that can give me an accurate high quality orthomosaic too.

Edited by Av8Chuck
Link to comment
Share on other sites

  • 3 weeks later...
On 2/27/2017 at 0:28 AM, Av8Chuck said:

 

I'm looking for software that can give me an accurate high quality orthomosaic too.

Accuracy over and above that of the UAS is going to be dependent upon your ground control points and their accuracy. With accurate GCPs, we are getting in the 1-1.5cm range on the surface for northings and eastings. You won't be able to meet that with a software-only solution. You are going to have to invest in a GPS unit that offers the accuracy levels you are hoping to achieve, or hire a mapping or survey firm to set control on your site that you can orthorectify your data to.

I know that the Inspire 1 can gather in the 30cm - 5m range of accuracy (roughly 11 inches to 16 feet). For real estate and some other application that is good enough. For construction and survey, it doesn't cut it.

Link to comment
Share on other sites

NADIR photogrammetry is mostly used for mapping, DEM, agriculture  or anything horizontal.

Oblique scanning of structures is used to gather data to create a 3D model of structures like cell towers, bridges, buildings, or anything vertical.  

Although applications like Pix4D and Acute3D appear to be similar, they are both used for photogrammetry, they are based on the application of difference science.  Pix4D is the application of photogrammetry for mapping and surveying.  Acute3D is the application of photogrammetry for creating 3D models.  

In order to build models that are accuarate enough for civil engineers and architects to use in the design process, you need to be able to blend different types of surveys, NADIR and obliques from the air but also obliques from the ground.  Applications that are developed for NADIR are not as effective at doing this as ones developed to create 3D models.  This isn't an issue of which is better, they're just different.

So the customers for oblique data are utilities: power substations, towers, wind turbines, power transmission lines etc..  Oil & Gas: pipeline, rig, refinery, environmental.  Communications: cell towers and switching buildings.  Municipalities: bridges, dams, aquifers.  Construction, infrastructure restoration, transportation, etc..

keep in mind that the ratio of modifying existing infrustructure and new construction is about 10:1.  Add to that drones can capture much more oblique data and about ten times faster than traditional survey methods, they can be a real game changer.  

 But the level of accuracy is critical for 3D models and as R Martin points out Inspires are no where near accurate enough.  

 

Link to comment
Share on other sites

Thanks for the list of possibilities and yes, having recently completed a college-level course in surveying I am well aware of the limitations of consumer grade UAVs for professional level surveying. Why would survey pros spend 70K+ on their birds if they could get the same precision with a Yuneec, right? I got to see a bird with integrated Trimble GNSS. Pretty sophisticated stuff.

I was more interested in understanding the why and hows of type of customers needing 3D models, and your list is appreciated.

Link to comment
Share on other sites

I'm looking for a method to place a 3d model into an android app and then have the drone fly around the 3d model and not have to rely on GPS to determine the position of the drone. A use case would be to conduct repeat inspections of a building. Complete a detailed 3d model once. Then use the 3d model to determine the position of the drone by comparing the video to what the drone sees while flying to what the drone would see flying around a 3d model.

I've contacted several 3d model companies but I have not found something yet. If anyone knows of such a product then please let me know.

Link to comment
Share on other sites

5 minutes ago, eColumbia99 said:

I'm looking for a method to place a 3d model into an android app and then have the drone fly around the 3d model and not have to rely on GPS to determine the position of the drone. A use case would be to conduct repeat inspections of a building. Complete a detailed 3d model once. Then use the 3d model to determine the position of the drone by comparing the video to what the drone sees while flying to what the drone would see flying around a 3d model.

I've contacted several 3d model companies but I have not found something yet. If anyone knows of such a product then please let me know.

There are several challenges to accomplishing what you want to do.  Without GPS there's no reference to the location of the building to the planet, there's also no scale or context between the 3D model and the real things being scanned, also we save every mission we fly, if its a successful mission we can repeat it without much difficulty.  

There's a lot of discussion about machine vision and SLAM but using it on a drone for sense and avoid is very different then using it for accurate scanning.  Its probably going to be a while before machine vision is integrated into workflows with enough accuracy that civil engineers can trust.  And its going to be expensive.  It will require much more than a hobby grade  IMU, we're testing the Applanix APX-15 (Trimble) just that IMU alone costs $18K.  Plus what it costs to integrate it with the other instrumentation so that you can collect accurate location data in a GPS denied environment will cost another $10K-$15K.

That's not to say that some toy drone manufacturer won't claim they can accomplish this today.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share