A pixel, or picture element, is the smallest item of information in a digital image. It’s that single square in the mosaic that colors the screen upon which you are reading these words.
It is essentially a dumb element, it knows only one thing — what color it is. It has no idea about the color of the other pixels around it. And, above all, it does not know what the totality of the pixels on the screen represents. A pixel is ignorant about whether it is playing a bit part in a digital photo of your dog or a completely black field.
In contrast to the pixel, the textual elements that you are reading here and markup language that tell your browser what to display are machine-readable, searchable and semantic. The future of the Internet is currently aimed at this semantic web, the web of organized data as described by Tim Berners-Lee, the indexed web of Google and the computational web of Wolfram Alfa.
This post, however, sings the praise of the dumb pixel.
Paradoxically, due to the paucity of information contained in the pixel, collections of pixels, aka images, can penetrate filters and can spread information through unexpected channels. A good example would be photographs of protesters in Iran carrying signs in English designed to be distributed through blogs, newspapers and social networks. The textual filters of the Iranian government cannot make sense of what the group of pixels is expressing.
Another interesting setting to observe the power of the pixel is in the emerging use of Twitter avatars as an auxiliary communication channel. Huh? Once again turning to the political events in Iran, many thousands of people have used a simple green overlay on their twavatars to signify solidarity with the protesters. A few of the more geeky users are inserting badges and messages in their twavatars.
Accompanying the 140 character tweet without consuming even 1 character is the meta-data, expressed in pixels: “This person supports the post-election protesters in Iran.”
The more geeky among you will point to advances being made in image recognition software. There will come a day that almost all images will be scanned, interpreted, amended with meta-data and converted into machine-understandable formats. OCR software is already quite sophisticated. On the other hand, there will be an endless game of whack-a-mole in which developers will try to construct images of dumb pixels that will fool recognition software. You can see this being played out today in the captcha images used in some site registration and blog commenting systems. Porn and spam merchants will, of course, lead the charge!
Until the day machines will be able to make as much sense of pixel mosaics as human beings can, let us take a moment to hail the humble pixel!
