[cv.jit.label] threshold smallest than & bigger than

Aug 19, 2008 at 10:19pm

[cv.jit.label] threshold smallest than & bigger than

Hiya Jitterlist!
From Russia with love again!

When I need to threshold small noise or other stuff that I don’t need to track it’s easy right? Yep! Just change minimum blob size in pixels by attribute [threshold].
But how to make also in the same time threshold “bigger than” objects?
Just I have situation where I know a needed size range!
For this time only one thing came into my head – unpack all data and compare area of blobs…but I use list “on the fly” and hope
that anyone of Jitter community have a beutirfull and simple idea!

Many thanks for your time!
Cheers

#39351
Aug 19, 2008 at 11:16pm

jit.op using min, max, < or >
using cv.jit object to filter out noise can be a smart move but also
consider cleaning the source before tracking.

On Wed, Aug 20, 2008 at 1:19 AM, outoff wrote:

>
> Hiya Jitterlist!
> >From Russia with love again!
>
> When I need to threshold small noise or other stuff that I don’t need to
> track it’s easy right? Yep! Just change minimum blob size in pixels by
> attribute [threshold].
> But how to make also in the same time threshold “bigger than” objects?
> Just I have situation where I know a needed size range!
> For this time only one thing came into my head – unpack all data and
> compare area of blobs…but I use list “on the fly” and hope
> that anyone of Jitter community have a beutirfull and simple idea!
>
> Many thanks for your time!
> Cheers
> –
> http://www.asyan.ru
>

#138359
Aug 20, 2008 at 12:01am

of cource I’ve use it before tracking, but when I’ve veery good matrix to find needed blobs I’ve also found some blobs that bigger than and I don’t want to track it. need only “middle” ;-)
may be I don’t understand all, sorry for this

#138360
Aug 20, 2008 at 12:41am

its ok, u can use jit.op operations after you track, the data is still a
matrix.

but working on lists (after jit.iter) can be easier.
i usually make it a list store in a coll and query mass, index etc.

On Wed, Aug 20, 2008 at 3:01 AM, outoff wrote:

>
> of cource I’ve use it before tracking, but when I’ve veery good matrix to
> find needed blobs I’ve also found some blobs that bigger than and I don’t
> want to track it. need only “middle” ;-)
> may be I don’t understand all, sorry for this
> –
> http://www.asyan.ru
>

#138361
Aug 20, 2008 at 2:19pm

That’s illustrate how I’ve suppose your idea, but not at all((
Hope you can find time to look at :)

Thanks!

#138362
Aug 20, 2008 at 8:50pm

And another test patch, bugs fixed, tested and works – calculate how much points in range, but how to remove these point coordinates from output list. Look at, it’s funy experiment ^_^
tnx!!!

#138363
Nov 8, 2008 at 5:07pm

Hello,
I’m currently using a Canopus ADVC-55 for video input into Jitter. The quality is farily low but so was the price.
I’m looking into either the “DFG13941e” or possibly the “Decklink SP” to hopefully improve quality.
Is the massive hit in data rate worth the quality increase? I mainly use Canon XL2′s or better for input thru a vision mixer.
I use a MacPro with 3×500 gig SATA2 Seagate drives setup with software RAID 0. Any thoughts?
Cheers,
Stuart

#138364
Nov 8, 2008 at 5:59pm

capturing and grabbing live data are two different beats, the DFG13941e will
actually be “cheaper” to process thne a dv stream, we are talking about
uncompressed firewire here.
and i envy your raid setup btw.

On Sat, Nov 8, 2008 at 7:07 PM, stuart smith wrote:

> Hello,
> I’m currently using a Canopus ADVC-55 for video input into Jitter. The
> quality is farily low but so was the price.
> I’m looking into either the “” or possibly the “Decklink SP” to hopefully
> improve quality.
> Is the massive hit in data rate worth the quality increase? I mainly use
> Canon XL2′s or better for input thru a vision mixer.
> I use a MacPro with 3×500 gig SATA2 Seagate drives setup with software RAID
> 0. Any thoughts?
> Cheers,
> Stuart
>
>
>

#138365
Nov 8, 2008 at 6:13pm

plus you can have more than one of them actively capturing at once on
a mac, which is not possible with standard firewire over DV.

On Nov 8, 2008, at 12:59 PM, yair reshef wrote:

>
> capturing and grabbing live data are two different beats, the
> DFG13941e will actually be “cheaper” to process thne a dv stream, we
> are talking about uncompressed firewire here.
> and i envy your raid setup btw.
>
> On Sat, Nov 8, 2008 at 7:07 PM, stuart smith
> jams.com> wrote:
> Hello,
> I’m currently using a Canopus ADVC-55 for video input into Jitter.
> The quality is farily low but so was the price.
> I’m looking into either the “” or possibly the “Decklink SP” to
> hopefully improve quality.
> Is the massive hit in data rate worth the quality increase? I
> mainly use Canon XL2′s or better for input thru a vision mixer.
> I use a MacPro with 3×500 gig SATA2 Seagate drives setup with
> software RAID 0. Any thoughts?
> Cheers,
> Stuart
>
>
>

#138366
Nov 8, 2008 at 7:30pm

Quality vs bandwidth? IS the IDG1394 ‘MUCH” better quality?
All the best,
Stuart

#138367
Nov 8, 2008 at 10:51pm

On Sat, Nov 8, 2008 at 9:30 PM, Stuart wrote:

>
> Quality vs bandwidth? IS the IDG1394 ‘MUCH” better quality?

it is up to a point, in the end both r just A2D convertors. the “shit in
shit out” rule applies here. you might get better answers if you explain a
bit what you want to do with your video. do you want to record, manipulate,
apply as texture, computer vision, buffer…

>
> All the best,
> Stuart
>

#138368
Nov 9, 2008 at 9:50pm

Hello,
Thanx for the replies.
I mainly capture video, manipulate and playback using multiple
videoplanes.
All the best,
Stuart

#138369

You must be logged in to reply to this topic.