YOUR ACCOUNT

Login or Register to post new topics or replies
Jason Rodriguez
Posts: 2
Hi,

Can you get access to per-pixel RGB values as a triplet array (so basically some sort of per-pixel seperated component output like pixelValue[x][y] would return a pixelColorValue[r][g][b])?
  Details E-Mail
Vladimir Golovin
Administrator
Posts: 3446
Filters: 55
Jason, there is an ExtractRGB component in channels, but it is not per-pixel -- internally, Filter Forge doesn't operate on pixels.

Could you please describe what you're trying to do? I had an idea about a Color Picker component that would allow extraction of color from a point at specific coordinates, but the idea wasn't finalized enough to be included into the beta.
  Details E-Mail
Mike Blackney

Posts: 375
Filters: 57
Vladimir wrote:
I had an idea about a Color Picker component that would allow extraction of color from a point at specific coordinates, but the idea wasn't finalized enough to be included into the beta.

It's not too hard to do ourselves anyway. The attached filter is one example smile;)

Pixel Color.ffxml
  Details E-Mail
byRo
an Englishman in Brazil

Posts: 138
Filters: 8
Drats, must get up earlier - Mike beat me to it!

Rô

RGB Picker.ffxml
_________________________________
My favourite question is "Why?".
My second favourite is "Why not?"
  Details E-Mail
Mike Blackney

Posts: 375
Filters: 57
Haha how funny that we both did pretty much the same thing smile8)

Yours is more useful, though - it actually lets you get the RGB like the OP asked for.
  Details E-Mail
byRo
an Englishman in Brazil

Posts: 138
Filters: 8
Quote
Mike Blackney wrote:
Haha how funny that we both did pretty much the same thing

Which makes it instructive to examine the differences. smile;)

- Moving the sample to centre, as you did, is a neater solution than moving the sampling point around.
- "Blending" using Profile Gradients is a neat trick that I'll have to remember.
- Didn't understand the width 1 impulse minus the width 0.5 impulse, though.

Rô
_________________________________
My favourite question is "Why?".
My second favourite is "Why not?"
  Details E-Mail
byRo
an Englishman in Brazil

Posts: 138
Filters: 8
Not sure if you're interested but, anyway, here's a trick to simplify the little square jump.

Rô

_________________________________
My favourite question is "Why?".
My second favourite is "Why not?"
  Details E-Mail
Mike Blackney

Posts: 375
Filters: 57
Quote
byRo wrote:
- "Blending" using Profile Gradients is a neat trick that I'll have to remember.


I only just worked that one out myself smile:) It's amazing how concise you can get some filters by cutting out the unnecessary nodes.

Quote
- Didn't understand the width 1 impulse minus the width 0.5 impulse, though.


Ahhh, change the position of the 0.5 impulse to -49.5 and the width of the 1 impulse to 2. That'll look more like what I wanted to get smile:D

Quote
Not sure if you're interested but, anyway, here's a trick to simplify the little square jump.


Much nicer!
  Details E-Mail
Jason Rodriguez
Posts: 2
I was wanting to get pixel values since this program looked like a great/easy way to test out different debayer algorithms (a Bayer CFA is the color grid that is on single-chip cameras).

Thanks for all the feedback.

I'm assuming I can take the color picker values from your graphs and then plug that as the values I would need to calculate the values for the demosaic algorithm . . . of course I would also need a FOR loop or something of that nature to get the picker to keep shifting down the image-not so sure that's possible . . . I haven't looked at the documentation yet, sorry if these are silly questions.
  Details E-Mail

Join Our Community!

Filter Forge has a thriving, vibrant, knowledgeable user community. Feel free to join us and have fun!

33,711 Registered Users
+18 new in 30 days!

153,533 Posts
+38 new in 30 days!

15,348 Topics
+73 new in year!

Create an Account

Online Users Last minute:

13 unregistered users.