Oscar Pizarro

Toward large-area mosaicing for underwater scientific applications

Severe attenuation and backscatter of light fundamentally limits our ability to image extended underwater scenes. Generating a composite view or mosaic from multiple overlapping images is usually the most practical and flexible way around this limitation. In this paper, we look at the general constraints associated with imaging from underwater vehicles for scientific applications - low overlap, nonuniform lighting, and unstructured motion

Large area 3D reconstructions from underwater surveys

and present a methodology for dealing with these constraints toward a solution of the problem of large-area global mosaicing. Our approach assumes that the extended scene is planar and determines the homographies for each image by estimating and compensating for radial distortion, topology estimation through feature-based pairwise image registration using a multiscale Harris interest point detector coupled with a feature descriptor based on Zernike moments, and global registration across all images based on the initial registration derived from the pairwise estimates. This approach is purely image based and does not assume that navigation data is available. We demonstrate the utility of our techniques using real data obtained using the Jason remotely operated vehicle (ROV) at an archaeological site covering hundreds of square meters.

Microbathymetric Mapping from Underwater Vehicles in the Deep Ocean

Robotic underwater vehicles can perform vast optical surveys of the ocean floor. Scientists value these surveys since optical images offer high levels of information and are easily interpreted by humans. Unfortunately the coverage of a single image is limited by absorption and backscatter while what is needed is an overall view of the survey area. Recent work on underwater mosaics assume planar scenes and are applicable only to situations without much relief. We present a complete and validated system for processing optical images acquired from an underwater robotic vehicle to form a 3D reconstruction of the ocean floor. Our approach is designed for the most general conditions of wide-baseline imagery (low overlap and presence of significant 3D structure) and scales to hundreds of images. We only assume a calibrated camera system and a vehicle with uncertain and possibly drifting pose information (e.g. a compass, depth sensor and a Doppler velocity log). Our approach is based on a combination of techniques from computer vision, photogrammetry and robotics. We use a local to global approach to structure from motion, aided by the navigation sensors on the vehicle to generate 3D submaps. These submaps are then placed in a common reference frame that is refined by matching overlapping submaps. The final stage of processing is a bundle adjustment that provides the 3D structure, camera poses and uncertainty estimates in a consistent reference frame. We present results with ground-truth for structure as well as results from an oceanographic survey over a coral reef covering an area of approximately one hundred square meters.

Advances in large-area photomosaicking underwater

In this paper we present a methodology for very high resolution acoustic mapping from an underwater vehicle. Based on the navigation, attitude, and bathymetric sensors that currently exist we show that the major issues for underwater mapping involve modeling sonar-specific artefacts due to beam pattern effects, calibration offsets of attitude sensors, and estimation and correction of tidal influences. We present an algorithm for modeling the beam pattern and for estimating calibration errors based upon specific vehicle maneuvers. Using a 675-kHz mechanically scanned pencil-beam sonar mounted on the Jason ROV, mapping results from an archaeological site at water depth 800 m in the Mediterranean are presented that are seen to be consistent and independently verifiable.

UWIT: underwater image toolbox for optical image processing and mosaicking in MATLAB

The propagation of visible light underwater suffers rapid attenuation and extreme scattering. This, in combination with the limited camera-to-light separation available on most imaging platforms, places severe limitations on our ability to optically image large areas of the sea floor at high resolution. We present a general framework for mosaicking large areas underwater with a specific emphasis on the issues that are unique to the underwater environment. At the individual image level, we examine the role of attenuation, scattering, and camera to light separation and present the tradeoffs involved in optimizing a particular imaging geometry. We also examine the arbitrary image-registration problem in the face of conditions prevalent underwater, namely a moving nonuniform lighting source and the effects of a featureless unstructured terrain. Our analysis is based on photomosaics encompassing several hundred images on archaeological, forensic, and geological expeditions from a diverse set of imaging platforms, including the NR-1 nuclear submarine, the manned submersible Alvin, the Argo towed vehicle, the Jason remotely operated vehicle, and the ABE autonomous underwater vehicle.

Advances in High Resolution Imaging from Underwater Vehicles

This paper shows results from our development of an extended MATLAB image processing toolbox, which implements some useful optical image processing and mosaicking algorithms found in the literature. We surveyed and selected algorithms from the field which showed promise in application to the underwater environment. We then extended these algorithms to explicitly deal with the unique constraints of underwater imagery in the building of our toolbox. As such, the algorithms implemented include: contrast limited adaptive histogram specification (CLAHS) to deal with the inherent nonuniform lighting in underwater imagery, Fourier based methods for scale, rotation, and translation recovery which provide robustness against dissimilar image regions, local normalized correlation for image registration to handle the low-feature, unstructured environment of the seafloor, multiresolution pyramidal blending of images to form a composite seamless mosaic without blurring or loss of detail near image borders. In this paper we highlight the mathematical formulation behind each of these algorithms using consistent notation and a unified framework. We depict some of our MATLAB toolbox results with an assortment of underwater imagery.

Towards image-based characterization of acoustic navigation

Large area mapping at high resolution underwater continues to be constrained by the mismatch between available navigation as compared to sensor accuracy. In this paper we present advances that exploit consistency and redundancy within local sensor measurements to build high resolution optical and acoustic maps that are a consistent representation of the environment.

In-situ attitude calibration for high resolution bathymetric surveys with underwater robotic vehicles

This paper examines the role of image-based navigation in the context of characterizing the standard long baseline navigation used by underwater vehicles for survey applications. Our work is based on looking at the displacement estimate that can be derived from registering overlapping imagery of the seafloor. Our approach is realistic in that it does not require large overlap and in that it can handle translational and rotational motions between image pairs in an unstructured terrain. We demonstrate our approach on a photographic survey conducted by the Argo towed vehicle covering several square kilometers off of Guam in the Pacific Ocean over a period of almost two months.

Imaging Coral I: Imaging Coral Habitats with the SeaBED AUV

In this paper we present a methodology for high resolution acoustic bathymetric mapping from a robotic underwater vehicle. Based on data obtained from navigation, attitude, and bathymetric sensors we show that precise calibration of attitude sensors is critical to obtaining high precision bathymetric surveys. We present an in-situ method for precision attitude sensor calibration based upon specific vehicle maneuvers. This method is demonstrated using data from an acoustic bathymetric survey of an archaeological site in the Mediterranean conducted by the authors with the Jason remotely operated vehicle.