Swarm Algorithms, Better Pictures, and Multicore Computers

Swarm Algorithms, Better Pictures, and Multicore Computers


I received an interesting email this morning, the Dr. Dobb's Report, a newsletter I subscribe to. The lead article is titled "Swarm Algorithms and (Better) Digital Photos." Whenever I see something like this, I am immediately reminded of the potential benefits of multicore processors on home computers. Surely, if there's a way to improve the quality of digital pictures, that's something everyone's going to want to be able to do.
But if improving the pictures involves complex algorithms, then those algorithms will have to be multithreaded, using Threading Building Blocks or some other technique, and the enhancement program will have to run on a multicore system; otherwise, users will be idly watching their screens, getting ever more impatient as they wait forever for the enhancement analysis to complete.
The Method: Particle Swarm Optimization Enhancement
Here's a snippet from the Dr. Dobb's Report message:
... researchers are turning to "swarm intelligence" as a means of enhancing digital photos. Swarm intelligence is an AI technique that's based on the collective behavior of decentralized, self-organized systems.
The technique ... uses a Particle Swarm Optimization (PSO) algorithm to enhance contrast and details without distortion.
Somewhat like Genetic Algorithms, PSO treats each version of an image as an individual member of a swarm and makes a single, small adjustment to contrast levels, edge sharpness, and other image parameters. The algorithm then determines whether the new members of the swarm are better or worse than the original, according to an objective fitness criterion.
The authors' paper, "Particle swarm optimization enhancement approach for improving image quality", was published in the International Journal of Innovative Computing and Applications in 2007, and is available for purchase.
Conclusion
Algorithmic techniques that are applicable to everyday activities are being actively researched. The advent of multicore computers and multithreading techniques like TBB present opportunities to bring these innovations to home and office desktop computers. I expect that, once people realize such software is possible, they'll want to have it. The beneficiaries will be the developers and software vendors who recognize the full potential of multicore processors, and bring these new products to market early on.