Processing Multibeam Imagery in SonarWiz

By David Finlayson, Chief Scientist

Multibeam echo sounder (MBES) imagery has improved dramatically in recent years and each new version of SonarWiz has introduced improved tooling for working with these data. Most of SonarWiz’s image processing and targeting tools are now also available for MBES systems. In this article we will take a look at the two different approaches offered by SonarWiz for processing MBES image data.

A MBES image differs from traditional sidescan (see Figure 1). Each beam in a MBES is roughly equivalent to a complete side scan channel. To create an MBES image, the portion of the beam time series that corresponds to the bottom detection (also called a “snippet”) is extracted and placed along the beam footprint. The snippets from successive beams are pieced together to form the image across the swath.

Fig 1 – Formation of a sonar image from (A) multibeam and (B) side scan sonars.
The multibeam image is built by piecing together the time signal (“snippets”) from each beam; the sidescan image is composed of a single time series. (after Lurton, 2020, figure 8.32)

The along-track resolution of MBES images is usually less fine than side scan sonars due to the wider beam apertures. While the quality of the reflectivity data is improved by the beam forming (less noise), the beam directivity patterns modulate the amplitude data making it more complicated to level the snippet intensity into a seamless image.

Finally, MBES are usually operated to optimize for bathymetry collection, so the systems maintain high angles of incidence with the seafloor. This makes it hard to detect microrelief that would be apparent in a side scan image collected close to the seabed at a high grazing angle. (Lurton, 2010)

SonarWiz currently supports two different workflows for processing MBES imagery: (1) convert the MBES imagery into a pseudo-side scan image; and (2) processing the intensity data on the bathymetric soundings. There are pros and cons to each approach.

Figure 2 – Waterfall views of MBES imagery.
Image constructed using (A) the Side Scan Engine at 4096 samples per channel, and
(B) the Bathymetry Engine at 512 samples per channel.

When you import MBES imagery using the side scan engine (See Figure 2A) SonarWiz can display up to 4096 samples per channel (8192 samples across the whole swath). The resulting imagery can be very high resolution if the multibeam packets support it (for example: S7K Snippets, ALL or KMALL Time Series, R2 TruePix, etc.). Once the data is converted to side scan, the operator can take full advantage of SonarWiz’s powerful side scan processing tools.

This workflow is a fast way to process hundreds of track lines and is recommended for target detection and identification since it offers the highest possible image resolution. However, the SonarWiz side scan engine does not raytrace the samples to the seafloor and cannot accurately represent overhangs or other complicated topography. Nor can it handle unconventional sonar offsets. For these situations, we recommend using the bathymetry engine.

The second MBES image processing option in SonarWiz is to import the data using the bathymetry engine (see Figure 2B). In this mode, SonarWiz assigns a single intensity value to each beam in the ping. The user can select to use the beam amplitude assigned by the manufacturer or SonarWiz can process the snippet intensity data and replace the beam amplitude with a statistically computed best value. In either case, the across-track image resolution is limited to the number of beams in the ping.

The advantage of this approach is that each image pixel is located precisely in X, Y and Z. So there is no ambiguity in the position of targets. Image mosaics created in this mode are tightly aligned. For many applications, the across-track resolution of shallow water multibeam systems is so high, that there are still tens of image pixels per meter which is more than enough to create beautiful seafloor mosaics. In addition, the bathymetry data makes it possible to see the same targets in both shaded relief and intensity waterfalls side-by-side which can be a powerful new dimension to target detection.

Figure 3 – Side-by-side waterfalls from the same bathymetry file.
(A) shaded relief bathymetry; and (B) beam amplitude intensity.

Of course, it is possible to use both methods in a single SonarWiz project. Users with both the side scan and bathymetry modules can select the processing method that makes the best use of their data. Either way, we have tried to make the processing of intensity data easy and efficient, with few steps needed to produce a clean image. 

Reference Lurton, X. (2010) An introduction to Underwater Acoustics: Principles and Applications. Second Edition. Springer, Heidelberg. pp. 362-363.