Descattering Nemo

The sun is bright, the water looks inviting, you’ve got your holiday snorkel and fins, your waterproof camera and you’re swimming out to that reef to see if you can snap a photo of Nemo. You catch a glimpse of orange with the telltale black and white stripes and click the shutter, a few shots of corals and clams, other little fish that dart in and out. Your friends and family will be so impressed with your social media updates when you get back to the hotel.

Unfortunately, all those shots you thought seemed so vibrant on the boat trip have a blue-green color cast over them, some of them are underexposed, others are overexposed, they all look a bit murky, if you’re honest, and although you thought you had found Nemo, he just looks like as much a common or pond-diving grey blur as the less bright fish you photographed for context.

No one is going to be impressed with this murky aquamarine gallery.

But, before giving up all hope, there is always the possibility of a photo editing package that can tweak the levels, boost the contrast, get rid of the oceanic color cast and perhaps even boost the saturation levels of the colors a little. The standard tools in common photo-editing software do go some way to fixing such problems with images. However, writing in the International Journal of Computational Science and Engineering, researchers from China, Japan, and Taiwan, have devised an approach to reconstructing an underwater scene that carries out color correction to overcome the blue-green hue of the underwater environment and “de-scattering” to remove the murkiness caused by particles in the water scattering so much of the light.

Huimin Lu, Yujie Li and Seiichi Serikawa of the Kyushu Institute of Technology in Kitakyushu, Japan, worked with Xin Li of Shanghai Jiaotong University, Jianru Li of Tongji University, China, and Kuan-Ching Li of Providence University, Taichung, Taiwan, to develop their marine photo fix. Of course, the system has the capacity to work with amateur holiday snaps, professional photos, and images recorded for research purposes equally well. Moreover, the system allows them not only to improve the image quality but to reconstruct a 3D, three dimensional, model of the scene that was photographed with reduced noise levels, better exposure of dark regions, and improved global contrast where the finest details and edges are often lost but can now be enhanced significantly.

Lu, H., Li, Y., Serikawa, S., Li, X., Li, J. and Li, K-C. (2016) ‘3D underwater scene reconstruction through descattering and color correction‘, Int. J. Computational Science and Engineering, Vol. 12, No. 4, pp.352-359.

Author: David Bradley

Award-winning, freelance science writer based in Cambridge, England.