Anyone wanna join? customize software

[xhacktly]

Hi folks, i am trying to start a project where the main goal will be to make a full visual and cheap 3d hologram and be able to send and transfer hologram over internet.

please see link under:

https://www.xcore.com/projects/led-cube ... d-software

Do u have other ideas?, thoughts?,facts?..
or relevant info, please join and discuss!:)

so to all coders, arduino nerds, chip-nerds and electronic freaks that wanna make this happen, please join:)

Do u have questions just ask, and we will discuss:D

[franksanderdo]

Hi & Hello,

are XCore beginners welcome?

I am still reading to get into it and I only will get some thing to play (XC1A and XK1A) beginning of June only.
If you can live with more questions then answers let me know.

See you around
Frank

[xhacktly]

Hi & Hello,

are XCore beginners welcome?

I am still reading to get into it and I only will get some thing to play (XC1A and XK1A) beginning of June only.
If you can live with more questions then answers let me know.

See you around
Frank

cool, yes i can live with that!;) the goal is to create something new! , i also have a lot of questions and are also a beginner.

i think the led cube part will be do able, but the 3d simulating software are still a nut to crack..
we have to research what kind of open source software that is out there, that we might can hack on, and it might be best to start with something simple..

example:

A one way rotating 4-square dice, that get filmed by a web-cam.

The software we write will simulate all the proportions/measurements of the dice (all sides of it, even the one the cam can not see, but we know the proportions ).
The web-cam will send this threw a network to an other client with the same software installed, and a led cube connected to a micro controller that is connected to the clients computer.
The signal will the be inverted, so the client will be able to see a full 3d hologram of the dice form the first client!

ideas?, questions?, bring it;)

[franksanderdo]

Hi & Hello,

Ideas? About a thousand :lol:

To get dimension of a peace of whatever ;) I believe we need either 2 cameras or a defined position of the peace to the camera. This makes the calculation of the dimension easier.

What I believe is the much bigger issue:
How to isolate the Object we want from the rest of the picture? I think there is some recognition software required!?
As I talk, sorry write, about that: It might be easier, instead of using a webcam, to use a uncompressed signal of a camera module. There might be some recognition software available already.

In parallel we can start thinking about the LED Cube.
I agree it should not be to difficult. just a question about the dimensions you have in mind.
From there we can decide how to "feed" it. I guess a set of shift Registers might be a good Idea!?
Starting point could be a single color cube which we can extend later to more colors.

What Do you think about this mixture of Ideas and questions?

See you around
Frank

[xhacktly]

Hi & Hello,

Ideas? About a thousand :lol:

To get dimension of a peace of whatever ;) I believe we need either 2 cameras or a defined position of the peace to the camera. This makes the calculation of the dimension easier.

What I believe is the much bigger issue:
How to isolate the Object we want from the rest of the picture? I think there is some recognition software required!?
As I talk, sorry write, about that: It might be easier, instead of using a webcam, to use a uncompressed signal of a camera module. There might be some recognition software available already.

In parallel we can start thinking about the LED Cube.
I agree it should not be to difficult. just a question about the dimensions you have in mind.
From there we can decide how to "feed" it. I guess a set of shift Registers might be a good Idea!?
Starting point could be a single color cube which we can extend later to more colors.

What Do you think about this mixture of Ideas and questions?

See you around
Frank

nice, yes good thinking!, yes i think we have to start small and later go for the ultimate goal:
"A full 3d hologram of a person or a hand,(in real time) or to think EVEN BIGGER(how cool would it be to have a hanging 3D led carpet i a room,that can visualize a full person in real time) (but of course this is the ultimate target/fantasy, and may something we're not able to do..so we should start with some easy first:P)


Good idea of having more web-cams working together!
this will be mush easier (maybe.:S)
with two cams, the person can have one at the front of the screen and one at the side, so we get two dimensions, and only have to replicate the third dimension.
(with 4 cams we will be able to get more measurements but the goal is to have as less web-cams as possible.)

"How to isolate the Object we want from the rest of the picture? I think there is some recognition software required!?"

I guess a set of shift Registers might be a good Idea!?

good question, i did not think about that, yes some recognition software would then required, what about using a (greencarpet/greenwall) to make this even easier?

a set of shift registers? (what is that ?)

i think this thread would be best to just discuss "how", and "ideas" and "goals" and do a lot of "research" and other relevant things in the stating process of the project,
then later we can create a thread for the "led cube" and one thread for the other topics (ex; recognition software, etc..) how do u think about that?

thanks for joining;) this project is going to be awesome!

[franksanderdo]

Hi & Hello,

about the threads here:
Yes I would say that we have to discuss first the Ideas a bit more global.
As soon as we start some detail on the issues we can separate the threats.

About the recognition of an object:
I believe that, with enough contrast, it should be possible to process an image frame in a way to distinguish lines.
If we work with 2 cameras in a defined angle and focus we "only" have to find the first line of our cube and then process from there.

To get an Idea how to get there I would like to start with one camera facing a 2d object and try to isolate it.
Let me have a look how we can easiest get a camera signal into the chip for processing.
I have an idea but have to see if this is a stupid idea first ;)

for the LED cube, I have seen projects on other Controllers. Maybe we can pinch there some Ideas?

See you around
Frank

[franksanderdo]

Hi & Hello,

just to drop the Idea that we can have a look at it later:

I did have a look in the kinect stuff from MS.
This might be an nice way of recognising shapes and dimensions?

See you around
Frank

[phalt]

Kinetic sounds like an interesting idea. I'm keeping a keen eye on where this project discussion is going ;)

[franksanderdo]

Hi Phalt,

do not keep an eye, discuss with us :lol:

OT
Is there a reason why I get logged out every few minutes?
It is a bit of work to log in each time again :roll:

See you around
Frank
P.S. it happened again during writing this post

[phalt]

There shouldn't be any reason for you being logged out - have you checked your browser settings regarding cache and cookies?

I just tested login a few times myself on various browsers and got no issues.