Finding spam in text- and HTML-based messages is possible thanks to a variety of detection and filtering mechanisms. However, when spammers embed their message text into images, detection takes on a new level of difficulty. Secure Computing said that it sees a growing trend towards this type of message delivery.

A spokesperson for the company said that its research shows a 200-percent increase in the amount of image-based spam; the company thinks such spam now accounts for approximately 30 percent of all junk mail. This points to an obvious problem, particularly when spam detection solutions don't account for this type of unwanted intrusion or fail to account for image manipulation possiblities.

"Traditional anti-spam software depends on content filtering techniques such as keyword filtering and Bayesian analysis to detect spam. Even the technology used to recognize characters from images, optical character recognition (OCR), is not effective on today's image spam," said Dr. Paul Judge, Secure Computing CTO. "Spammers are using advanced mathematical and graphical techniques like random modification of image pixels and dynamic construction of images from multiple components to bypass spam filtering tools."

Secure Computing said that it works to battle against image-based spam using a reputation-based approach. The company collects statistical data on email senders and the types of email they send. To accomplish that Secure Computing uses more than 7000 sensors located in 48 countries. Accuracy in determining the potential for a message to be spam increases as the sensors collect more data over time. The company said this type of approach has proven effective to help its solutions eliminate image spam before it reaches company mail servers.