Connecting 3D/360° Tour Buyers to Service Providers powered by
the Most In-Depth, Relevant and Up-to-Date Content for the 3D/360° Community
Join FREE! 50+ Benefits
Should I buy a Camera? Quiz How Much to Start 3D Tour Business? Calculate How Much Can I Make as a Pro? Calculate Win More Listings? Calculate
Option #1
Find a 3D Tour Photographer Map
We Get Around Network Find a Pro Service
Matterport Service Provider? Join We Get Around Network
Option #2
Book a Matterport Pro in 60 Seconds
We Get Around Network Instant Booking Service
Powered byKoaWare
Matterport Service Provider (USA)? Join KoaWare Free! Register Now
Last 30 Days: 59,421 Page Views | 13,074 Visitors | 96 New Members
WGAN Knowledge Base | Total Posts: 58,660  |  Total Topics: 8,955
4,995 WGAN Members in 125 Countries
We Get Around Network Forum
WGAN Training Academy (Learn more)
WGAN-TV Channel
Matterport Workshop 3.0 Training for Newbies
Part 5 with Matterport Marketing Mgr. Amir Frank
Matterport Training Academy
60+ Hours | Succeed Faster
Matterport Scanning | Matterport Workshop 3.0
Private Message @DanSmigrod for free ...
Subscribe Now WGAN-TV Program Schedule | Need Help? | All Videos
GeoCVKonami

Konami Code found in GeoCV Viewer4314

3rd Party
Service
Metroplex360 private msg quote post Address this user
Quote:
Originally Posted by AntonYakubenko
Quote:
Originally Posted by Metroplex360

15x higher res = 61440 x 30,720.


I consider resolution to be the number of pixels, like when we say a camera has 12 megapixels resolution. When you multiply both width and height by 15 you get 225 higher resolution


@AntonYakubenko - Oh my gosh -- I am so embarrassed now that I read that. [facepalm]. Of course, you are correct.

Quote:
Originally Posted by AntonYakubenko
So the assumption was that Matterport panoramas are 6 images 1024x1024 each, which gives us 6 megapixels resolution. The support of 2k tiles (and thus 6x2048x2048 resolution) doesn't really improve the real resolution, only the formal one, which is seen from you example images in another thread - the difference between 1k and 2k tiles is very minor.


I do believe that it improves quite a bit -- and I am grateful that you referenced by previous post. There's a lot more clarity in the 2k tiles -- however, certainly on a more powerful camera, 2k tiles would have more data in them -- I'm certain of that.

Quote:
Originally Posted by AntonYakubenko
We produce panoramas of around 14000x7000 resolution (98 megapixels), which can then be cut in different resolution tiles up to 4k per tile. 98/6=16.3, which is close to our 15x estimate. Subtracting nadir and zenith do not change this estimate, as we don't capture them as well for now


Makes perfect sense -- thank you for clarifying.

What it sounds like to me is that GeoCV is interested in providing a high level of zoom -- or when displayed on a faster computer with a fast Internet connection at a high resolution -- the panoramas will appear extremely high resolution.

--

Quote:
Originally Posted by AntonYakubenko
Anyway, it's not real 8192x4096 resolution as your images have shown.


Would you mind helping me out with this -- I understand what you are saying, but I lack the correct terminology.
Post 26 IP   flag post
3rd Party
Service
Metroplex360 private msg quote post Address this user
Quote:
Originally Posted by AntonYakubenko
Our solution to be released later down the road won't require a DSLR camera and a GigaPan motorized panoramic head. With a total investment cost from you as a service provider of around $1000 you'll be able to get ~100 megapixels panoramas as a part of a whole 3D virtual tours solution. Until we release our solution to the most interested beta testers you could just believe us, or not.


One struggle that you will find throughout our forum is that we invested in Matterport cameras in order to provide a service that is of value to our clients.

The high price of the Matterport Camera actually HELPS to prevent oversaturation in the market and many of us have not been too excited about the decrease in price as it is a double-edged knife.

If GeoCV succeeds with a $1,000 solution -- it has the potential to hurt our industry.

Quote:
Originally Posted by AntonYakubenko
We've developed a computer vision algorithm that automatically aligns 360 panoramas to the 3D scan. 3D computer vision is the area of our expertise, we've been working in this industry for over a decade.


That is -extremely- cool and I believe that it satisfies my inquiry that I made on the other thread concerning patents.
Post 27 IP   flag post
3rd Party
Service
JuMP private msg quote post Address this user
@AntonYakubenko
It is really cool that you can align Pano to 3D mesh even they were taken seperately.

I don't know how to measure the real resolution of a photo. And I googled it with no result either.
Would you please tell the way? I want to learn and would like try to measure them. Thank you!

There is another reason why Matterport pano looks not so sharp. It is the size of the picture and compress rate limited by the network bandwidth and loading time.
The pano I pasted is only 1MB. The compress rate is 4.17%.
And compress rate of your DEMO pano is 9.55%.
Post 28 IP   flag post
GeoCV
Co-Founder
AntonYakubenko private msg quote post Address this user
Quote:
Originally Posted by Metroplex360
Quote:
Originally Posted by AntonYakubenko
Anyway, it's not real 8192x4096 resolution as your images have shown.


Would you mind helping me out with this -- I understand what you are saying, but I lack the correct terminology.

Well, imagine you have a super crisp and detailed 1 megapixel image (1000x1000). Let's resize it 10x on each side to 100 megapixels image (10000x10000). The formal resolution of the image we've got is 100 megapixels, but the real one is 1 megapixel.

There are different ways how you can measure the real resolution. When you read some consumer cameras reviews you can see the testing guys shooting and analyzing special linear patterns. You won't see thin lines on small resolution images, they would just become a gray mess. There are more scientific and accurate ways as well.

And here is one approximate approach to measuring the real resolution when you don't have a special pattern in the image. Imagine you have 1000x1000 pixels image. Let's resize it to 800x800 and then resize the resulting image back to 1000x1000. Compare the processed image with the original one. Do you see a loss of details? Try to look at the real details, not the noise.

No difference? Let's go further. Resize 1000x1000 to 600x600 and then back to 1000x1000, compare with the original one. If you see that there is a visible loss of real details it means that the real resolution of our 1000x1000 image is somewhere between 600x600 and 800x800. If you make more steps in such a process, for example, adding 700x700 step, you make a more accurate estimate.

Please mind that this is a very approximate way and whether there is a substantial loss of details or not is subjective. But if could give a decent approximation for any image without any special patterns required. I'd also notice that "real resolution" could depend on the JPEG quality, camera parameters during shooting, the lighting in the scene, etc. So to compare apples to apples you need more "laboratory" testing setup. But this simple approach described above could give a good estimate of the real resolution for any image.
Post 29 IP   flag post
GeoCV
Co-Founder
AntonYakubenko private msg quote post Address this user
Quote:
Originally Posted by Metroplex360
One struggle that you will find throughout our forum is that we invested in Matterport cameras in order to provide a service that is of value to our clients.

The high price of the Matterport Camera actually HELPS to prevent oversaturation in the market and many of us have not been too excited about the decrease in price as it is a double-edged knife.

If GeoCV succeeds with a $1,000 solution -- it has the potential to hurt our industry.

I totally understand. Consider our solution as an opportunity, not a threat. An opportunity to do more business and scale it further. Matterport says 2.5% homes for sale in the US have its 3D tour. I hope together we can successfully go after other 97.5%.

I don't think your main competitors are other photographers who bought a more affordable 3D tours solution. And in most cases neither are realtors who often don't want to shoot by themselves. By providing a more affordable service using our solution and reaching a larger audience you can do grow your business.

Imagine the world when at least 20% of realtors doing 80% transactions have a habit of ordering a 3D virtual tour, professional photos and a floor plan. 80%, not 2.5%. Would you have a better business then? Maybe more people in the team, and more cameras, which don't cost you a fortune. The elasticity of the market is subject to testing, but if we look on the industry from the outside - everyone (home buyers, home sellers, agents, brokerages, photographers, listing portals) would benefit if more and more listings would have a high-quality 3D tour and professional photos.

We're here to make this future come true together. Too pathetic, I know, but that's the truth.
Post 30 IP   flag post


3rd Party
Service
Metroplex360 private msg quote post Address this user
Quote:
Originally Posted by AntonYakubenko
I don't think your main competitors are other photographers who bought a more affordable 3D tours solution.


They aren't today because no solution matches Matterport's offerings. RealVision and iGuide are two solutions that offer more affordable pricepoints and are very successful in certain markets to cost conscious Realtors.

If one throws GeoCV into the mix wherein a new batch of $1,000 camera owning photographers leap onto the field it could do the following if successful:

#1) Force Matterport Owners to match lower prices, thus lowering overall income.
#2) Move to the GeoCV platform in order to match lower prices.

This is assuming that at the time that GeoCV is commercially available that Matterport will still exist in a relatively "same state" as it does today. For all we know, Matterport may be selling a $1,000 camera before your launch too and this problem will not be limited to GeoCV's entry in the marketplace.

Quote:
Originally Posted by AntonYakubenko
And in most cases neither are realtors who often don't want to shoot by themselves. By providing a more affordable service using our solution and reaching a larger audience you can do grow your business.


Possibly. I'd love to hear some other opinions on this as I could argue for or against this idea.

Quote:
Originally Posted by AntonYakubenko
Imagine the world when at least 20% of realtors doing 80% transactions have a habit of ordering a 3D virtual tour, professional photos and a floor plan. 80%, not 2.5%. Would you have a better business then?


It would be a different world. Again, I would love to hear some opinions from others on this because there's a potential Pros & Cons list a mile long.

I think from the customer's perspective spending less money is a positive. We serve the customer, so ultimately we must bend to their needs in order to truly be a service-based industry.

Quote:
Originally Posted by AntonYakubenko
Maybe more people in the team, and more cameras, which don't cost you a fortune.


Do you have some fuzzy stats that you have considered? Does anyone on the forum want to speculate how a $1000 system with the promise of ... 1/2 the time for a scan (?) ... would change their business model?

i.e.
1 Day -- 2 Matterport Scans for $600
1 Day -- 4 GeoCV Scans for $600.

What would be the potential pros/cons? Hopes of booking more lower priced tours for steadier business? Or a con being managing more clients, billing, etc -- more ideal for agencies that an individual service provider?

Again, I'd love to hear other people's thoughts on a what if scenario. The same scenarios would apply to a 'What if Matterport released a new camera that was 2x faster?'
Post 31 IP   flag post
LeventeSolczi private msg quote post Address this user
I would like to know when and where be available the geocv technology.
Post 32 IP   flag post
59368 32 7
destitute