Ultrasonic Imaging Improvement by a Co-Operative Approach

The goal of this paper is the construction of an accurate scene model to be used by an underwater vehicle inspect and navigate in the environment.To this end, a co operative approach aimed at improving the luallty of short-range acoustic images created by a beamforming process is proposed, which is mainly -selui for human understanding and robotic applications.Our approach is essentially a voting method: it consists in the integration of two methods devoted to the estimation of the angle of arrival for each time nstant and the estimation of the time of arrival for each xed direction.Each sampled time instant of each beam signal contributes with an opportune quantity on a voting space.At the end of the process, maxima recovery in the voting space and post-processing methods are applied to recover the precise silhouette of a scene.Actual linear-scan sonar data and angular-scan simulated sonar data were used to verify the validity of the method.Results and error measures are reported in order to compare our method with commonly used methods.showing advantages and drawbacks of the co-operative approach.