Fourier-based registrations for two-dimensional forward-looking sonar image mosaicing

Abstract

This paper presents a method to build large-scale mosaics adapted to underwater sonar imagery. By assuming a simplified imaging model, we propose to address the registrations between images using Fourier-based methods which, unlike feature-based methods, prove well suited to handle the characteristics of forward-looking sonar images, such as low resolution, noise, occlusions and moving shadows. The registration between spatially and temporally distant images resulting from loop-closing situations or registrations in featureless areas are feasible, overcoming the main difficulties of feature-based methods. The problem is cast as a pose-based graph optimization, taking into account the uncertainties of the pairwise registrations and being able to incorporate navigation information. After the optimization, a consistent mosaic from different tracklines is generated with increased resolution and higher signal-to-noise ratio than the original images, while the vehicle motion in x,y and heading is also estimated. © 2012 IEEE.This work has been partially supported by the FP7-ICT-2011-7 project PANDORA (Ref 288273) funded by the European Commission, the Raimon project CTM2011-29691-C02-02 and the grant BES-2008-00609

Forward-looking sonar image mosaicking by feature tracking

“…The third class of methods calculate the translation and rotation parameters in the Fourier domain. For example, in 2012, Hurtós et al [8] applied the Fourier-Mellin transformation [9] to image registration for forward-looking sonars. It should be noted that only when the images were high-pass filtered, the normalized cross spectrum matrix has a clear peak response [3].…”

Real-time mosaicing with two-dimensional forward-looking sonar

“…While our previous efforts [5][6] concentrated on offline generation of acoustic mosaics, this paper tackles the problem of real-time FLS mosaicing with the aim of providing an online map of the areas that are being inspected. By achieving real-time capabilities, the advantages and applications of FLS mosaicing are further expanded.…”

“…The method is able to achieve real-time performance by using a dedicated FPGA. In our previous work [5][6], we have tackled the problem of FLS mosaicing by utilizing Fourier-based registrations. We have evaluated the performance of this registration against traditional featurebased techniques [10] as well as against the method used in [7], concluding that Fourier-based registration provides better results for aligning FLS images [6].…”

“…We take advantage of a Fourier-based registration methodology and a pose-graph formulation to perform a global alignment over the vehicle odometry and estimated constraints between sonar frames. The advantages of using a Fourier-based registration are threefold: first, it is able to cope well with the challenging characteristics of the FLS imagery (low resolution, low SNR and intensity alterations due to the vantage point) [5], second, it offers better performance when conducting registrations in featureless areas that lack prominent features [10], and third, it has a reduced computational time due to the fast implementation of the FFT (Fast Fourier Transform). The pipeline of the proposed algorithm deals with every incoming frame, determines links with previous frames in the mosaic, and optimizes the global position of all frames while still maintaining real-time requirements.…”

An Adaptive Image-stitching Algorithm for an Underwater Monitoring System

“…Several modified phase correlation algorithms based on the Fourier transform have been proposed in the literature [5][6][7]. The algorithms enable the measurement of translation, rotation and scaling factors between two images.…”