There is estimated to be over 420 million pornographic pages on
the internet with thousands more being created every day. With the
advent of Web 2.0 the problem is becoming increasingly difficult to
manage as the dynamic nature of the internet is making traditional
web filtering techniques less effective. Internet video alone will
account for 57 percent of all consumer Internet traffic by 2014
with a large proportion of that being pornographic.
The images on a pornographic webpage are the one constant which is
both region and language independent. Therefore to effectively
filter pornographic content it is essential to analyze the image
and video content on the page along with the other elements.
Image Analyzer's high performance multithreaded architecture makes
inline image analysis a feasible solution for web filtering
vendors. The optimum deployment of the technology is as part of a
blended solution working in conjunction with traditional filtering
techniques such as URL databases and lexical analysis.