Forums > Jitter

Chopping up images from a live webpage

October 31, 2010 | 10:41 am

Hi,

I’ve been trying a number of solutions but nothing is really working for me here. I have a live webpage and I want to chop up the images on it into individual matrices. I have tried jweb to jit.scissors but jweb doesn’t output a matrix. I have also tried jit.desktop but it is way to cumbersome and slow (especially for my laptop screen real estate). There are over 200 images so downloading each to a jit.qt.movie doesn’t seem very practical. It seems like it would make more sense to just cut up the webpage if that’s possible.

Any other suggestions for objects or externals I should look into?


October 31, 2010 | 12:46 pm

You said downloading would be a hassle but I don’t really see another way to do it. Depending on how your webpage is programmed you could use [jit.uldl] and [jit.str.regexp] to find the exact image URL to download. What webpage are you using?

lh


October 31, 2010 | 12:58 pm

It’s my own page, hosted locally – i have over 200 images aligned in a grid and they are ALL rapidly changing (every 100ms or so) through the page’s JavaScripting. I have tried to metro a "download url" to [jit.uldl] whereby all images go to a folder. My problem is that the cpu lags while downloading and reading the images every 50 ms from the folder. I thought that by programming a javascript for a webpage, it could do the image retrieval freeing up Jitter’s resources to do other things. This is why I want Jitter to "view" the entire page and break it into columns and rows for analysis. Does that make sense?


October 31, 2010 | 12:58 pm

It’s my own page, hosted locally – i have over 200 images aligned in a grid and they are ALL rapidly changing (every 100ms or so) through the page’s JavaScripting. I have tried to metro a "download url" to [jit.uldl] whereby all images go to a folder. My problem is that the cpu lags while downloading and reading the images every 50 ms from the folder. I thought that by programming a javascript for a webpage, it could do the image retrieval freeing up Jitter’s resources to do other things. This is why I want Jitter to "view" the entire page and break it into columns and rows for analysis. Does that make sense?


October 31, 2010 | 1:06 pm

You don’t say if you need to do this programmatically (that is, based on a request that won’t be known until runtime), or not.

If no — there is a plug-in for the Firefox browser called FlashGot. Install it, navigate to your page, then select "FlashGot All" — it pulls down all the images on the page to your hard drive.

Kurt


October 31, 2010 | 1:22 pm

Haven’t tried Flashgot, but will check it out now. The problem isn’t getting the images to the HD – uldl with an indexed array will do that fine – it’s that the images change every 100ms, so I need to stream them – then analyze each one in Jitter. Each one is only 16kb so it’s not terribly intensive when run locally.


November 1, 2010 | 2:06 pm

I figure that is I hook up my laptop to a larger, secondary display then jit.desktop might work. It is really not the most elegant solution though.


November 1, 2010 | 7:27 pm

I don’t fully understand how jit.desktop will help.
If it is hosted locally then you probably have a folder full of images on your hard drive, yeah? In this case all you need is a umenu that is auto-loaded with the contents of that folder (see the "prefix" message/attr) that sends "read" messages to a jit.qt.movie object. If that is not a solution, please let us know what exactly you are having trouble with (display of images?, downloading from server?, outputting to screen?, etc.)


November 1, 2010 | 8:06 pm

Yes, that is correct. I have the images in a folder. The site is hosted locally but the images are downloaded every 100ms as they change. So I can access them locally or online. And populating a umenu isn’t a problem.

So how might I get all the images from the umenu to appear in a grid simultaneously? Should I be exploring some javacripting?


Viewing 9 posts - 1 through 9 (of 9 total)