Day 36: SSS Image Processing – Masterpiece Edition | River Warrior

🖼️ DAY 36: SSS IMAGE PROCESSING

⏱️ Estimated Reading Time: 15 Minutes | 🎓 Level: Professional Hydrographer / Sonar Analyst

From Raw Waterfall to Georeferenced Mosaic – Enhancing Side Scan Sonar Imagery

Instructor: Engr. Rokib Hossain | River Warrior Academy

🏠 Course Homepage

1. Why Side Scan Image Processing Matters

Raw side scan sonar data is displayed as a “waterfall” of acoustic intensity vs distance. However, the raw image suffers from slant range distortion, variable gain, and navigation errors. Processing turns this raw data into a georeferenced mosaic that can be used for:

Detecting wrecks, pipelines, cables, and debris.
Mapping seabed texture and sediment types.
Inspecting scour holes and riprap.
Providing evidence for legal cases (e.g., lost cargo).

Without proper processing, the image may be misleading or unusable.

🧠 Golden Rule: Always process side scan data with the same careful calibration as bathymetry. Slant range correction is mandatory for accurate target positioning.

🌊 River Warrior Pro-Tip: Jamuna Ferry Wreck

In 2023, we used side scan to locate a sunken ferry in the Jamuna River. Raw data showed a faint shadow. After applying TVG and slant range correction, the wreck’s outline became crystal clear – and we guided the recovery team directly to it.

2. Understanding the Waterfall Display

The waterfall display (echogram) scrolls vertically, with time/distance increasing downward. Each horizontal line represents one ping, showing acoustic intensity across the port and starboard swath. Dark tones (low intensity) represent shadows; bright tones (high intensity) represent strong echoes (hard seabed, rocks, metal).

Waterfall display: dark shadows, bright hard returns, and neutral seabed.

3. Slant Range Correction (Angle to Ground)

Side scan measures slant range (direct distance from towfish to target). To get the true horizontal distance, you must apply slant range correction using the depth of the towfish and the beam angle:

Ground range = √(slant² – (towfish height)²)

Correcting to ground range removes the “stretching” of near‑range pixels, making the seabed appear true to scale. Most processing software (SonarWiz, Hypack, Qimera) does this automatically when you enter the towfish depth profile.

✅ Without slant range correction, targets appear artificially compressed close to nadir and stretched at far range – leading to wrong position estimates.

4. Gain, Contrast & TVG Adjustments

Side scan data often has variable intensity across the swath due to beam pattern and range. Three key adjustments:

TVG (Time‑Varied Gain): Amplifies distant echoes to compensate for signal spreading. Essential for uniform image.
Overall gain: Controls the brightness of the entire image.
Contrast / histogram stretch: Enhances subtle features by spreading the intensity range.

Start with manufacturer’s default TVG, then fine‑tune.

🎛️ Interactive Gain Simulator

Adjust the slider to see how gain affects a simulated side scan line (one ping).

Gain (0 = low, 1 = high): 0.50

Simulated side scan ping: higher gain reveals more noise but also weak targets.

5. Georeferencing & Layback Correction

To place side scan data on a map, you need:

Vessel GNSS position (lat, lon) with heading.
Layback correction: Horizontal offset from GNSS antenna to towfish.
Towfish depth to compute layback (see Day 23).

Processing software (e.g., SonarWiz) uses these to create a georeferenced .xtf or GeoTIFF. Ensure layback is correctly entered; otherwise, targets will be misplaced by metres.

📌 Always test georeferencing by comparing a known target (e.g., a buoy or wreck) with its charted position.

6. Mosaicking: Stitching Lines Together

A mosaic combines multiple side scan lines into a single continuous image. Steps:

Georeference each line (apply layback, slant range correction).
Export each line as a georeferenced raster (e.g., GeoTIFF).
Blend overlapping areas using feathering or weighted average.
Output final mosaic with a colour palette (grey or false colour).

Mosaics reveal large‑scale features (pipeline routes, scour depressions, wreck fields).

Multiple overlapping side scan lines merged into a seamless mosaic.

7. Target Identification & Classification

Common targets on side scan sonar:

Wrecks: Distinct high‑intensity returns (metal) with a long acoustic shadow.
Pipelines & cables: Linear bright line, often with a shadow on one side.
Rock outcrops: Irregular shape, high intensity, shadow depending on relief.
Scour holes: Dark depression with a bright rim.
Marine vegetation / sand ripples: Fine‑scale texture, no strong shadow.

Always measure target dimensions using the georeferenced image. Document with screenshots and position.

🔍 In the Jamuna River, we identified a series of buried pipelines using side scan: the exposed sections appeared as bright linear features, while buried sections were invisible – critical for dredging safety.

8. Case Study: Jamuna Wreck Detection & Recovery (2024)

Scenario: A 30‑m barge sank near the Jamuna Bridge. The owner requested location and recovery.

Method: Side scan sonar (450 kHz) with a towfish at 5 m depth. Survey grid with 40 m line spacing.
Processing: Slant range correction, TVG, mosaic, and georeferencing using SonarWiz.
Result: Mosaic clearly showed the barge on its side, 22 m long, with a 15 m shadow. Coordinates delivered to divers.
Lesson: Without slant range correction, the wreck would have appeared 10 m shorter – recovery team would have searched the wrong area.

📸 Always export a high‑resolution PNG of the mosaic with scale bar and north arrow for the client.

9. Side Scan Image Processing Checklist

Raw .xtf (or .s7k) files logged and backed up.
Towfish depth profile entered (for slant range correction).
Layback and heading corrections applied (Day 23).
TVG and gain adjusted to produce a balanced image.
Slant range correction applied (convert to ground range).
Georeferenced images exported per line (GeoTIFF).
Mosaic created with overlapping lines blended.
Targets digitised and measured (position, length, orientation).
Final mosaic exported with scale bar, legend, and metadata.

Click items to track progress (saved in browser).

10. Software & Tools

.htmlHypack Side Scan那样Real‑time waterfall, mosaicking, target marking那样hypack.com.htmlQPS Qimera (Side Scan module)那样Georeferencing, mosaic, waterfall display那样qps.nl/qimera.htmlTeledyne PDS (Side Scan)那样Integrated with SeaBat systems那样Teledyne.htmlOpen source: Sidescan Toolbox (MATLAB)那样Basic processing for researchers那样GitHub search

Software	Processing Features	Link
SonarWiz (Chesapeake Tech)那样Mosaicking, slant range, target digitising, export那样SonarWiz

11. Frequently Asked Questions

What is the difference between slant range and ground range?

Slant range is the direct distance from towfish to target; ground range is the horizontal distance. Slant range must be corrected to ground range for accurate scaling.

Why do I need layback correction?

The towfish is behind the vessel (typically 30‑80 m). Without layback, the target position would be incorrectly placed ahead of the actual location.

Can I process side scan data without dedicated software?

Some open‑source tools exist (e.g., Sidescan Toolbox for MATLAB), but for production work, SonarWiz or Hypack is strongly recommended.

What resolution mosaic should I aim for?

For wreck/pipeline detection, 0.25‑0.5 m per pixel is typical. For large area reconnaissance, 1‑2 m is sufficient.

How do I interpret a dark shadow?

A shadow indicates an area where the acoustic pulse is blocked (e.g., a wreck, rock, or steep slope). Measure shadow length to estimate object height.

12. Action Items & Next Steps

📌 Obtain a sample .xtf file (many manufacturers provide demo data).
📌 Load it into SonarWiz (or free viewer) and adjust gain/TVG.
📌 Practise identifying a target: measure its length and shadow.
📌 Proceed to Day 37: Sediment Classification & Mapping.

🖼️ Day 36: Side Scan Sonar (SSS) Image Processing and Mosaicking