Colour-Consistent Structure-from-Motion Models Using Underwater Imagery
Mitch Bryson · Matthew Johnson-Roberson · Oscar R. Pizarro · Stefan B. Williams
TL;DR
Using 3D structure from structure-from-motion, the method corrects underwater attenuation and lighting effects to produce colour-consistent photo-textured models from AUV imagery.
BRIEF
Water absorbs and scatters light in ways that change an object's apparent colour depending on how far the camera is from it. Lighting patterns and vignetting add further distortions. Together these make underwater images hard to interpret and hard to classify automatically. This work uses the 3D structure recovered by structure-from-motion to estimate the camera-to-surface distance for each pixel. It then corrects for distance-based attenuation, vignetting and lighting pattern, which makes the colours of photo-textured 3D models more consistent. The method is demonstrated on imagery from an autonomous underwater vehicle in two environments.
Abstract
This paper presents an automated approach to correcting for colour inconsistency in underwater images collected from multiple perspectives during the construction of 3D structure-from-motion models. When capturing images underwater, the water column imposes several effects on images that are negligible in air such as colour-dependant attenuation and lighting patterns. These effects cause problems for human interpretation of images and also confound computer-based techniques for clustering and classification. Our approach exploits the 3D structure of the scene generated using structure-from-motion and photogrammetry techniques accounting for distance-based attenuation, vignetting and lighting pattern, and improves the consistency of photo-textured 3D models. Results are presented using imagery collected in two different underwater environments using an Autonomous Underwater Vehicle (AUV).