Side-scan sonar image matching

A conceptual definition of a holonic processing framework to support the design of information fusion systems

“…To perform the association between images-based and elevationbased objects, DTM (from IFC u) and mosaic (from IFC s) act as global guides (H Al , H As , H Tr ). Several images-based objects or elevation-based objects may correspond to the same sea-bottom object, providing a more detailed representation of such an object due the multiple observations according to several viewpoints (H Es , H Co , H Pr ) [95]. Multi-views classification or characterization algorithms may be advantageously involved for this task [93].…”

Data fusion of multi-beam sonar and side-scan sonar base on feature contour registration

“…Reference [6] and [7] adopt Chamfer registration method to acquire MBES combination image and the SSS image, but this method may lead to registration parameters inconsistent in the part of image mosaic. In 1998, Daniel S adopted a pair of goals and shadows to register images, and his method can also apply to register the homologous sonar image, which hasn't been slant range corrected [8] . In the same year, Zu-xun Zhang et al, aimed at regular remote sensing images, proposed a different sensor, different resolution remote sensing image, fast automatic registration method [9] .…”

A feature-matching method for side-scan sonar images based on nonlinear scale space

“…There have been a number of methods used for sonar image matching, including using the highlight area and shadow zone of sonar images [3]; however, this only works for images with no orientation or scale change. Rominger used the theory of belief functions to obtain the optimal transformation in the registration progress [4]; however, this incurs a significant computational expense.…”