Visual analysis software that was originally created in 2000 has recently expanded to include the detection of pornography within digital data transmission. The detection method and software have been in place for some time, but the recent release of Image Analyzer 3.0 offers significant improvements that will allow its practical use in future applications in real time.

The company behind the technology, Image Analyzer Ltd., was contacted in 2000 by a major pharmaceutical company to produce software which would aid in dispensing the correct amount of medication into receptacles.

Since the accuracy of the system was critical to ensure each patient’s health, the Image Analyzer team created a visual analysis system based upon artificial intelligence; after some time, the project expanded into correctly identifying human bodies inside of images, and later was directed towards images with a pornographic nature.

Today, the company works closely with ISPs, managed service providers and software developers to ensure that their users are properly abiding by the terms and conditions that were set forth.

Pune, a city located in the western Indian state of Maharashtra, has already adopted this technology in their recently opened system, Pune Unwired, which offers a broadband wireless connection to all its citizens. Partnered with Image Analyser, the Pune Municipal District can detect visual pornographic content on hard drives and data transmission on PCs, PDAs, mobile phones, and off all things, even iPods.

The technology is quite complex, using multiple algorithmic-based modules to examine images, in real time, initially on skin color; the engine is able to distinguish between different regional skin colors as well.

The system then uses yet another module to analyze edges, curvature, and body size in the images; which, allows it to determine whether the image is potentially pornographic based on probability.

The use of this technology is certainly catching on quick; with corporations and manufacturers integrating these modules into their systems, it won’t be long before you see widespread adoption on social networking sites (e.g., MySpace,  Facebook), bundled with child monitoring software for the Internet, and quite possibly used in corporations to ensure employees aren’t viewing objectionable material.

While the system doesn’t detect pornography 100% of the time, it has a very high success rate, and it only takes one detection for the system to notify the proper authorities.  

Related:

Entry was posted on Tuesday, April 10th, 2007 at 8:29 am under Adult content, Software.  Print This Post

Read more entries by George Gardner.
Stumble It!