Hi! I'm Bubo.
Show me a detail and I'll tell you
where I've seen the full image before.
I memorize your images. I don't need metadata.
I can find all your images solely by their content.
Get started with HTTP REST. We even prepared a Ruby gem and Heroku add-on for you!
Imprint: spratpix GmbH & Co. KG, Am Fährberg 25, 24107 Quarnbek, Germany
Bubo is the cute little owl you found on the top of this page, the avatar for our new optical search technology. We chose this name because owls are known for their excellent memory and vision.
Conventional search technologies use so-called metadata, which are textual descriptions stored alongside the image. Sadly, this additional content is easily lost while modifying or converting image files.
Without metadata, conventional text-based search technologies are unable to find the image. And even then, results can only be identified if the user somehow guessed one of the keywords used inside the metadata. Conventional search masks are therefore optimized to auto-complete the users search terms to match stored keywords.
Our new search technology Bubo is fundamentally different. No textual content is required to retrieve an image, because we’re analyzing the actual graphical content. This means we can positively identify a JPEG image even after it’s been scaled, modified and saved as a GIF. If this sounds like bigheaded bragging to you, feel free to look at the attached examples.
Our search algorithm not only captures the overall impression but additionally stores every detail on its own. This enables our users to search for a composite to retrieve each used original image. Such a level of precision has previously been too slow and too expensive. Our enabling algorithm effectively reduces this calculation time to mere seconds.
It's fun. It's free. Be bold and try it out now!
If you're a hobbyist and just interested in the technology, write to hajo(at)spratpix(dot)com and tell us about it!
Hajo just loves discussing technology with smart people like you.