Human eye inspires advance in computer vision (w/Video)

Jun 18, 2009

Inspired by the behavior of the human eye, Boston College computer scientists have developed a technique that lets computers see objects as fleeting as a butterfly or tropical fish with nearly double the accuracy and 10 times the speed of earlier methods.

The linear solution to one of the most vexing challenges to advancing has direct applications in the fields of action and object recognition, surveillance, wide-base stereo microscopy and three-dimensional shape reconstruction, according to the researchers, who will report on their advance at the upcoming annual IEEE meeting on computer vision.

This video is not supported by your browser at this time.
Inspired by the behavior of the human eye, Boston College computer scientists have developed a technique that lets computers see objects as fleeting as a butterfly or tropical fish with nearly double the accuracy and 10 times the speed of earlier methods. BC computer scientists Hao Jiang and Stella X. Yu, who developed a novel solution of linear algorithms to streamline the computer's work, will present the team's findings at the IEEE Conference on Computer Vision and Pattern Recognition 2009, which takes place June 20-25 in Miami. Credit: Hao Jiang, Boston College

BC computer scientists Hao Jiang and Stella X. Yu developed a novel solution of linear algorithms to streamline the computer's work. Previously, computer visualization relied on software that captured the live image then hunted through millions of possible object configurations to find a match. Further compounding the challenge, even more images needed to be searched as objects moved, altering scale and orientation.

Rather than combing through the image bank - a time- and memory-consuming computing task - Jiang and Yu turned to the mechanics of the human eye to give computers better vision.

"When the searches for an object it looks globally for the rough location, size and orientation of the object. Then it zeros in on the details," said Jiang, an assistant professor of computer science. "Our method behaves in a similar fashion, using a linear approximation to explore the search space globally and quickly; then it works to identify the moving object by frequently updating trust search regions."

Trust search regions act as visual touchstones the computer returns to again and again. Jiang and Yu's solution focuses on the mathematically-generated template of an image, which looks like a constellation when lines are drawn to connect the stars. Using the researchers' new algorithms, computer software identifies an object using the template of a trust search region. The program then adjusts the trust search regions as the object moves and finds its mathematical matches, relaying that shifting image to a memory bank or a computer screen to record or display the object.

Jiang says using linear approximation in a sequence of trust regions enables the new program to maintain spatial consistency as an object moves and reduces the number of variables that need to be optimized from several million to just a few hundred. That increased the speed of image matching 10 times over compared with previous methods, he said.

The researchers tested the software on a variety of images and videos - from a butterfly to a stuffed Teddy Bear - and report achieving a 95 percent detection rate at a fraction of the complexity. Previous so-called "greedy" methods of search and match achieved a detection rate of approximately 50 percent, Jiang said.

Source: Boston College (news : web)

Explore further: Scientists track Internet usage as it pulses across the globe daily (w/ Video)

add to favorites email to friend print save as pdf

Related Stories

Search technique for images recognises visual patterns

Mar 16, 2005

Dutch researcher Mirela Tanase has developed a new technique for finding images using search engines. Her technique is based on how the human eye recognises objects. It can increase the success rate of certain search operations ...

Computer vision may not be as good as thought

Jan 25, 2008

For years, scientists have been trying to teach computers how to see like humans, and recent research has seemed to show computers making progress in recognizing visual objects. A new MIT study, however, cautions ...

Attention grabbers snatch lion's share of visual memory

Aug 07, 2008

Our visual memory is not as good as we may think, according to research funded by the Wellcome Trust – but it can be used more flexibly than scientists previously thought. In a study published today in the ...

Researchers Give Computers Common Sense

Oct 17, 2007

Using a little-known Google Labs widget, computer scientists from UC San Diego and UCLA have brought common sense to an automated image labeling system. The common sense comes as the ability to use context ...

Out of sight, out of mind? Not really

Aug 23, 2005

By playing a trick on the brain, neuroscientists at MIT's McGovern Institute for Brain Research have discovered one way that humans naturally recognize objects.

Recommended for you

Apple sees iCloud attacks; China hack reported

27 minutes ago

Apple said Tuesday its iCloud server has been the target of "intermittent" attacks, hours after a security blog said Chinese authorities had been trying to hack into the system.

HP supercomputer at NREL garners top honor

3 hours ago

A supercomputer created by Hewlett-Packard (HP) and the Energy Department's National Renewable Energy Laboratory (NREL) that uses warm water to cool its servers, and then re-uses that water to heat its building, has been ...

Turner channels removed from Dish amid pact spat

3 hours ago

Channels such as Cartoon Network and CNN are no longer part of Dish's programming lineup as a deadline has passed for the satellite TV provider and Turner Broadcasting to renew their distribution agreement.

User comments : 1

Adjust slider to filter visible comments by rank

Display comments: newest first

jimbo92107
not rated yet Jun 18, 2009
Call it a "recognition matrix" or something. Trust region is a goofy term that requires translation into more normal language...like recognition matrix or recognition point set.