An electron microscope image of a fly eye

Credit: Albert Einstein College of Medicine, Yeshiva University

Related articles

• Single-pixel camera creates hi-res images
• Photo-harvesting software creates 3-D world
• Robots hunt for woodpeckers
• Air guitar you can wear
• Bionic eye helps patients see the light

SYDNEY, 29 August 2006: The humble fly on the wall has inspired a new digital video technology capable of resolving detail where none could be captured before, paving the way for better surveillance cameras.

"When it comes to seeing, even a tiny insect brain can outperform any current artificial system," says Russell Brinkworth of the University of Adelaide in South Australia. "As we can, they can see detail in light and dark at the same time."

By mimicking the way insects see, Brinkworth can produce digital video which records minute detail in bright light and deep shadow simultaneously. The technique solves a significant problem for surveillance cameras, where lighting can be irregular and clarity is critical.

Traditional cameras use a single average light setting to control the brightness of an image. This works well, says Brinkworth, if there are similar levels of lighting across an entire scene. "But if the lighting is uneven or if somebody's deliberately trying to fool [the camera] by creating shadows, then it doesn't work well at all."

So Brinkworth looked to biology for a solution.

In nature, the individual cells of a fly's eye adjust to parts of an image independently, in order to capture the maximum amount of information about a scene. So while a traditional camera adjusts to the overall light intensity of a scene, Brinkworth's digital video technology adjusts each pixel separately.

"This means that even in difficult lighting conditions, such as a person standing in front of a window, you can see both the person's face and the scenery outside at the same time," says Brinkworth. "Something a traditional camera cannot do."

Brinkworth used microelectrodes to measure signals sent via individual neurons in a fly's brain in response to flashing lights or moving images. He then mimicked this mechanism using computer software and hardware.

While his software can be used to enhance detail in existing video footage, Brinkworth's goal is to introduce his technology at the capture stage. "What we want to do is to wire this into existing camera sensor technology – our software would be written to a computer chip that would sit between the sensor and the digital converter."

But Brinkworth emphasises that this is only the beginning. The study of insect visual systems has opened his eyes to a suite of other possibilities, including video cameras that can detect moving objects, compress and transmit video at incredible speed, and detect and measure the speed of very small objects moving in the distance. Not surprisingly, Brinkworth's team has already received support from the U.S. Air Force.

"Nature's a good model because it works and it's efficient," says Brinkworth. "But it's a fundamentally new way of thinking about vision technology."