Method for producing an image of a submerged object (3), e.g. a shipwreck or the sea bottom, comprising the steps of emitting acoustic waves from a first transducer toward a first chosen volume (2), receiving acoustic waves reflected from said volume, the receiver (4) including a two-dimensional array of acoustic sensors being capable of measuring the phase and amplitude of the received waves, from the measured phase and amplitudes calculating the direction of the reflected waves, detecting the maximum amplitude or amplitudes of the waves received from each angular segment and the distance to the reflecting object, the combination of directions of the reflected waves and distances providing a three-dimensional model of said volume, emitting acoustic waves toward a second volume (6) at least partially overlapping said first volume, and repeating steps described above, detecting at least one common feature (3) in the overlapping parts (7) of the measured volume (2, 6) and combining said models by adjusting the relative position(s) and size(s) of said feature(s) for generating a three-dimensional image based on the combination of said models.