Thanks to mathematics, the quality of underwater photographs is about to get dramatically better. The first-of-its-kind method for radically improving underwater shots combines a mathematical formula or algorithm with a filter normally used for land photography.
"This is a brand new solution to solving the problem of underwater image degradation," notes lead researcher Professor Yoav Schechner of the Faculty of Electrical Engineering at the Technion - Israel Institute of Technology.
"Underwater photos tend to be degraded and lacking in clear details because of the 'veiling' effect of ambient light," explains Schechner. "From above the water's surface, ambient light scatters into the line of sight -- an effect commonly known as 'backscatter.' This causes poor visibility in even the clearest water." He adds that the veiling effect increases with distance, making many details undetectable.
Schechner and graduate student Nir Karpel realized that photographic images could be greatly improved by eliminating the backscatter effect. The two then developed an algorithm that -- combined with a polarizing filter readily available for $20 to $100 -- compensates for backscatter.
Once the photographs are taken using the filter, they are transferred to a computer program, which then applies the algorithm, compensating for the underwater image degradation caused by backscatter. Unlike standard photo imaging programs that treat a photographic image as a whole, the new method corrects different elements -- such as objects that are closer or distant -- individually, according to need.
According to the researchers, this process also provides estimates of underwater distances, resulting in improved picture quality and photographs with three-dimensional depth.
Schechner says the program could be placed on a chip embedded within the camera or be integrated with existing video systems to vastly improve underwater video quality. In the future, it could be utilized for a myriad of applications including engineering, marine biology and archeology, underwater mapping, recreational photography and underwater rescue. The researchers are negotiating commercialization of the program.
The findings by researchers at the Technion-Israel Institute of Technology were presented at the Institute of Electrical & Electronics Engineers (IEEE) Conference on Computer Vision and Pattern Recognition in Washington, DC in June.
Published by Globes [online] - www.globes.co.il - on Thursday, September 02, 2004