[X]Assistant

[Ksalador]

Hello to all!

I am looking for a developer / programmer who knows and understands well whether sensors on the front panel of the iPhone X and Xs screen measure space and how it is measured.
The task is to create an application that, based on sensors on the front, can measure the space and convert it into figures (centimeters, inches, feet, etc.). It is important to understand whether it is possible to create a volume, how well it is possible to create this volume and translate it all into a figure (prototype of the space scanner, but based on the front panel)

[t.r.o.n.]

Ksalador
What of what you say is not implemented in the standard ios 12 application - roulette? (the question is not trivial, but for apple license points)
The front camera is always quite flawed due to the focal length. Therefore, relying on something serious with her is not realistic.
Everything related to such tasks is implemented in the standard library ios sdk - ARKit

Post has been editedt.r.o.n. - 22.11.18, 12:07

[Ksalador]

1. If I needed a roulette, I would download a roulette.
2. It is necessary to write the program under the front-end. Need a space scanner. So that a person does certain things in space and can look at the phone screen what he needs to do (instruction).
3. It is good that this is in the standard library, therefore I am looking for a person who knows this library and we will be able to talk with him in more detail in order to implement the product.

[t.r.o.n.]

Ksalador @ 11/23/18, 1:00 AM

3. It's good that this is in the standard library.

Therefore, he said that frontal flawed. Count on real use - it is impossible.

[Ksalador]

t.r.o.n. @ 11/23/18 09:18

Count on real use - it is impossible.

Why not? What is wrong with them? I am sure that they are made not only to use Face ID, there are a lot of other applications, such as separating a person's face from the main background and so on.
You say that you really can not use, bring the argument what their inferiority?

[t.r.o.n.]

Ksalador
For face recognition, you need a minimum resolution. But this has nothing to do with it.
The front camera is initially oriented to work at a distance of no more than a meter + a large angle. For publication in the trash is more than enough. For recognition - all the more.
But the vehicle is talking about the analysis of space. It is for this task that the camera is flawed. Well, you can not get a little bit of real information from it.
You think just like that the built-in roulette is focused on a normal camera.

[Ksalador]

t.r.o.n. @ 11/23/18 11:30

Roulette is focused on a normal camera

Roulette works with that camera, because it is convenient to watch and measure =)
Ok, how can we test with you the hypothesis that this camera is flawed and will not be able to measure correctly at a distance of more than 1 meter?

[t.r.o.n.]

Ksalador @ 11/23/18 11:33

because it is convenient to watch and measure =)

For the sake of Ponte, they would use everything, if there was such an opportunity.
They would add some kind of crap to pick up points or the like.

Ksalador
Well, as if I did not need to check. Played already with arkitom. Yes ... And I did not say that the camera does not measure more than a meter. I say that with the help of the front-end, in principle, you cannot measure it correctly ..
You want to walk on other people's rakes - your right. What can I do here?

Post has been editedt.r.o.n. - 23.11.18, 11:39

[Ksalador]

I want to see this rake. Where to read about what was tested, what kind of real functionality did these cameras have?
I do not want to spend time and money on the development of what can not be done or unrealistic.
For me it is now important to confirm or refute the hypothesis that the front sensors on the iPhone X and Xs can measure the space qualitatively and convert the visible object into a figure with all sizes. I need facts to prove or disprove this hypothesis. What you wrote is also a hypothesis that is not confirmed by the facts (see here how the camera sees and here it measures this space and so it can only measure to here, but these cameras can measure the space like this and there are other data shows us).
No offense will be said, but your words sound like this: The earth is flat, I have already rode the world and saw the ends of the earth.
My acquaintances wrote that it was realistic to do it, but no one either confirmed or refuted it to me. It turns out you need to either take and write software, or look for information about what is real or not real.
Therefore, I ask for help, let me read who has experienced and preferably put all this in with pictures.
Someone once said that using microsoft kinect it is impossible to measure the space and they were wrong, measured and very well, I saw software at exhibitions that it measures.

[Azathtot]

Ksalador @ 11/23/18 05:55

For me it is now important to confirm or refute the hypothesis that the front sensors on the iPhone X and Xs can measure the space qualitatively and convert the visible object into a figure with all sizes. I need facts to prove or disprove this hypothesis.

and the facts are very simple. They are still the ancient mathematicians expressed. That measurement accuracy is directly proportional to the rangefinder base.

Ksalador @ 11/23/18 05:55

Someone once said that using microsoft kinect can not measure the space

You compare the value of the base of the link and the value of the base of the phone.

[Ksalador]

Friends, the hypothesis is considered confirmed or not, when there is a fact. There is no fact and this is a fact. I do not compare the size of Microsoft kinect and the phone, since none of you will tell how many people have Microsoft kinect and how many phones. You really can not count the phones, but only assume. But how this problem is solved or not, you can answer. If you take and check how it counts. Therefore, the question is open, looking for a person who is ready to check it together.

[t.r.o.n.]

Ksalador
Call things by their proper names.
You personally have no facts. They are among others and based on them, knowledge and experience, they give you recommendations. It turns out that the process is more important to you than the outcome.

From experience (no offense), such criteria are important only for the divorce of investors. The developer does not need this.

[Ksalador]

Good.
Suppose made sure. A new task, we can measure an object using an AR kit with a conventional camera and a phone supporting technology. To know exactly the volume and length of the arc.

Post has been editedKsalador - 24.11.18, 02:27

Reason for editing: Errors in words

[Azathtot]

Ksalador @ 11/23/18, 20:26

New task, measure object ball

It depends on the size of the ball, the distance to the ball, and the desired error.

[Ksalador]

[email protected], 12:20

It depends on the size of the ball, the distance to the ball, and the desired error.

I can say maximum that I am interested in the error of 1 mm. The distance is no more than 1 meter. The size of the ball, and maybe not the ball, may be different. How can this issue solve ARCore now?

[Azathtot]

Ksalador,
none

[Ksalador]

AzathtotUm, while the roulette application, which does not need to be calibrated, works ...

[Azathtot]

Ksalador
With a certain accuracy, above 1mm declared by you.

[Ksalador]

AzathtotWhere to read about this accuracy?

[Azathtot]

Ksalador
You start the roulette, measure the distance and it clearly shows you the error. Better than +/- 1/2 inch I have not seen