When you enter some stores, you're not just being watched -- you're being mined.

An emerging technology called video mining is allowing stores and retail chains to use surveillance camera footage to study customers and produce accurate consumer data.

Video mining uses computer software that sorts through video footage to determine things such as how many and what types of people walk through a certain area of the store or stop to look at a certain product, said Rajeev Sharma, associate professor of computer science and engineering.

Using normal store surveillance cameras, the software sifts through the video feed for information or cues chosen by the store and then produces empirical data, in almost real-time, without a human ever seeing the video, he added.

Sharma is the chief executive officer of Advanced Interfaces, Inc., 403 S. Allen St., a company that offers video mining as one of its services.

This technology has already proven invaluable in better understanding customers and consumer trends, Sharma said.

"What do people do in stores? Everything is very impersonal, and our technology can monitor where people really go and watch how they really react to certain products," he said.

Sharma said the technology is even being developed to study peoples' facial expressions to determine reactions to certain products a store is selling. It can also be used to distinguish between male and female shoppers' reactions.

Another application may be for a grocery store, where this technology could be used to monitor checkout aisles.

The software can be programmed to watch the number of people at each checkout and can alert the manager when another aisle needs to be opened, Sharma said.

Raj Acharya, head of the department of computer science and engineering, said the data is collected in terms of classification and, in most cases, the software has proven to be better than humans at determining shoppers' gender just by their faces.

Ralph Oliva, professor of marketing and a member of the Advanced Interfaces advisory board, said current video mining technology benefits both vendors and customers.

Video mining allows stores to sell products more efficiently. Fewer resources are wasted on products that don't sell, so the products offered can be of higher value, and consumers can find better quality products for the money they spend, he said.

A driving force behind this technology is the fact that businesses want to be better prepared to serve their customers, Oliva said.

"Smart businesses are always anxious to gain more insight into consumer wants and needs, and this is a very innovative approach to doing just that," he said.

Consumers are getting pickier, and video mining technology allows for faster and better alignment of the market toward the needs of the customer, Oliva said.

For instance, if a store manager decides to display a certain product, he could determine almost immediately whether shoppers responded positively to it, as long as enough shoppers had passed by the display.

Nirmal Pal, executive director of the eBusiness research center at Penn State, explained in an e-mail message that this technology is essential in the information age.

"Customer needs and preferences are constantly changing in this digital age," he said. "Customers have many choices and information available to them at their fingertips."

Stores must then be able to gain new insight quickly, Pal said.

"Providers of products and services need real-time feedback so that they can respond quickly to the ever-changing needs and preferences of their customers," he said.

Acharya likened the use of this technology to what is done in Wal-Mart stores, where the layout is designed in such a way that all products seem to be in just the "right" place.

Sharma said that when this software is designed, countless examples are programmed into the computer so it can learn to recognize emotions from certain expressions or gestures, much like how we learn to read facial expressions.

For the most part, current video mining applications are fairly simple. Questions such as how long a customer spent looking at a particular display or product label, or how many customers passed a given area at a given time, can be answered easily using the data generated. However, this technology is becoming more advanced and more widespread, Sharma said.

He added that the technology is much more accurate than having a person with a paper and pencil keeping customer data and is also much safer because humans only see the final data and do not study individual customers.

If stores can guarantee the video footage is being studied only by the software and not by people, Acharya said, then most customers would not be concerned about their privacy and would not mind being studied.

There may be privacy issues associated with this technology, but they are no greater than when someone hands their credit or debit card to a store employee, Pal explained.

This technology can also be applied to any field where monitoring is used, such as watching over infants in nurseries or elderly in hospitals, Acharya said.

The uses of this technology will only grow as time goes on, Sharma said.

"It will be very widespread -- 29 million cameras are being used now, helping with security, and obviously it will grow," he said. "We want to make public areas safe, but also efficient and friendly, and we can add that intelligence."