Computer Programmers Borrow Eye Technology
Computer processing of video images may become twice as accurate with 10 times the speed of earlier models, thanks to what scientists are imitating in the human eye. “The linear solution to one of the most vexing challenges to advancing computer vision has direct applications in the fields of action and object recognition, surveillance, wide-base stereo microscopy and three-dimensional shape reconstruction,” a report in Science Daily said.
A team at Boston College team noticed how the eye performs a rough global search, then zeroes in for the details. They devised their software after this linear algorithm. It avoids having to comb through haystacks of data for the needle of interest. “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,” said Hao Jiang of the research team.
The technique allows 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.” It increased their detection rate to 95% over the 50% rate of earlier methods.
OK, so who wrote the software in the human eye that these guys reverse-engineered?