Future of Website Images - $200 To Whomever Can Implement Solution
I stumbled across this link the other day, and immediately I was impressed. Just watch the video. This is the future. Grab a hold or be warned! Best. Idea. Ever. I will personally pay for an implementation of this paper.
Basically, this method allows you to resize an image to any dimension, large or small without really any noticeable loss of detail (in most cases). I think this is really a huge step forward in web images. The comments on the above link keep referring to how great it would be to have in photoshop. Bah, people who use photoshop can already do these effects with the stamping tool, but there is no method for scaling images properly on the web as of yet. Adobe just hired the guy, and I would hazard to guess he is not headed for the photoshop team, but for the Flash team. Imagine flash implemented images that would scale wide banners across the page, whether at 800×600 resolution or 1680×1050! No more worrying about resolution issues when designing around image intensive sites.
The solution is fairly easy to implement. You can find the PDF of the whitepaper (warning 20+mB PDF across a slow connection).
You can also find an actual java demo of it here. It doesn’t handle image growing, but it is a fairly trivial matter to modify it to allow you to do so. Oh, and I do not see that he has released the sourcecode, but if you are really that interested in the source, you can most likely download the JAR and use JAD to take a peek (I do not condone such behaviour) . Check local laws regarding such things before you do it. It is all pretty well spelled out in the paper, so I would certainly check there first.
I will personally pay $200 to whomever can implement this in flash/actionscript with the following requirements:
- Ability to take in any image, and preprocess the required ’seams’ in a datafeed
- Implement the project in flash actionscript
- Allow for dynamic resizing (grow/shrink) of images/flash files in a browser (on events from browser resize)
- Provide me with the source for such a project, and rights to use it as I will
- ability to dynamically shrink/grow source image as per document spec.
The resulting swf must be able to detect browser resizing and apply the preprocessed calculations to resize the image on brower resizing. I am honestly unsure if flash files can detect brower resize, but I am going to assume that it can. Feel free to use the code elsewhere, I am not asking for an exclusive license, just that you provide me with the code and license to do as I please with it. I guess it comes down to a knowledge of flex/flash and the willingness to implement an already fully described algorithm. $200 isn’t much, but it is the most I can offer right now. It as least some incentive ;). It really could be the next great advancement to the web. Or it could be another whitepaper relegated to the confines of academia. Still, watch the video and let me know what you think.
I’m really going to doubt that anyone would implement it for that amount, but it’s worth a shot, eh?
Anyway, the idea is beautiful. Let me know what you think as far as the potential of the app goes. Submissions should be made to short_ship [at] yahoo.com for consideration for the reward. All valid replies must be made before end of September 2007. Hell, even if it does get fully licensed to the world via GPL or otherwise, the reward still remains the same.
Even if you cannot implement the solution, think of the possibilities of such a project. You could design your site around any resolution, and the graphics will remain completely scalable. One could post images without worry of scaling height/width of result image and still have a properly scaled image to whatever the target resolution is. It is a big step for image development. I’d love to see a fileformat built around this spec! Flash is the obvious choice as there is a dynamic element, but I’m sure there could be ajax solution/server side detection solutions built in to server proper images.
Just. imagine.
Even if you don’t write it for me. It is still damn impressive technology. Simple theory applied to a unique problem.
Feel free to disagree ![]()
Bookmark & Share
Related Posts
- Fixing Transparent Images In Internet Explorer
- 10 Firefox Extensions For Website Owners
- CES 08: Part 4 - The future of LCDs and Plasmas
- Google Clears Up Images Near Adsense Ads
- Improvements To Blogware Blogs: Suggestions Please!
- Speedup Your Website With HTML Optimizer Pro
- Replace Gravatars With MyBlogLog Images





Comment by Ryan Wagner on 6 September 2007:
It’s already been done in Flash:
http://swieskowski.net/carve/
Comment by Kline on 6 September 2007:
Ryan. Yes it has been ‘done in flash’ but its not useful or serve any real functional purpose. just a simple tech demo. It’s not tied to screen sizing, and it doesn’t grow.
Points I made sure to bold in my post
I saw the flash demo awhile ago, but couldn’t find the link when I wrote the article. That guy is 80% there, but it just really doesn’t show off the use (or the use I see for it) of the algorithm.
And he doesn’t give out his code. He’s got 2 out of the 5 bullets going for him
Comment by Kline on 6 September 2007:
Anyway, the point isn’t supposed to be me mooching for software development. $200 is pretty much nothing at all. Of course, it’s probably only 100-200 lines of actionscript, and hell, I know a limited amount of AS, but not enough to really get by when it comes to image manipulation.
I guess I’d rather hear thoughts on whether I was just completely drunk last night, or if this seriously has some potential for design. Flash image wrapper that will allow images to be resized arbitrarily to any dimensions should make web design much easier, and would allow for design to be done regardless of screen/browser size, dimensions, whatever. So a site will look different on 640×480 than it does on my 22″ widescreen, instead of designing for 640×480 and it being just a tiny vertical stripe of content in an otherwise empty wasted space.
Thats how I envision the techonology being used.