Enhancing Utility Detection with Clean GPR Data Using Background Subtraction Filters

Ground Penetrating Radar (GPR) is an invaluable tool for detecting underground utilities, but it often captures a variety of signals beyond the utilities themselves, including unwanted noise from soil layers and other flat-lying reflectors.

These distractions can make it difficult to clearly identify the true targets, such as pipes or cables, that appear as hyperbolas on the radar scan. One of the most effective ways to remove this “rubbish data” and improve clarity is through the use of a background subtraction filter. 

In this article, we’ll explain how a background subtraction filter works, why it is essential for utility locating, and the step-by-step process of using this filter to improve GPR data.

What is a Background Subtraction Filter?

A background subtraction filter is a signal processing tool used in GPR to remove or eliminate horizontal noise and flat-lying reflectors, which are often not related to the utilities or subsurface objects of interest. These horizontal signals or unwanted signals may originate from:

  • Soil boundaries or layers of earth with different properties.
  • Flat-lying geological features like bedrock, water tables, or other naturally occurring formations.
  • Surface reflections from nearby structures, roads, or even the GPR equipment itself.

When GPR is used to detect underground utilities, the most important features are usually the hyperbolas, which represent subsurface objects like pipes, cables, or other utilities. These objects cause GPR signals to reflect in a distinct hyperbolic pattern because of the shape and orientation of the buried object relative to the GPR unit.

By applying a background subtraction filter, GPR operators can remove the unwanted horizontal signals, or “background noise,” allowing the hyperbolas to stand out more clearly.

Why is Background Subtraction Important for Utility Detection?

In GPR surveys, especially in urban environments or complex geological settings, a significant amount of horizontal noise can obscure the true reflections from underground utilities. This can make it challenging to differentiate between irrelevant subsurface features and the actual targets.

The background subtraction filter is essential because:

  1. It Removes Distractions: By eliminating horizontal reflectors and flat-lying noise, the filter helps to focus the data on vertical and point-like features such as pipes and cables. This makes the hyperbolas stand out, which represent utilities, more visible.
  2. Enhances Hyperbola Visibility: GPR is most effective when it can clearly detect hyperbolic reflections. These indicate objects like pipes or cables. The background subtraction filter emphasises these hyperbolas by removing unnecessary data, improving the operator’s ability to locate and map utilities accurately.
  3. Improves Interpretation Efficiency: With less noise in the data, GPR operators can more quickly and easily interpret the results. This leads to more efficient surveys and reduces the likelihood of missing important utilities or making errors in utility detection.

Step-by-Step Guide on How to Apply Background Filter to your GPR Data

Applying a background subtraction filter is a standard process in most GPR data processing software. Here’s a step-by-step guide on how to apply this filter to your GPR data:

Step 1: Collect GPR Data

Before applying any filters, collect your GPR data first by scanning the area of interest. Ensure that your GPR unit is calibrated for the specific soil conditions and that you are using the appropriate frequency for the depth and size of the utilities you want to detect.

Step 2: Review the Raw GPR Data

Once the data is collected, examine or review the raw GPR scan. In this initial view, you may notice significant amounts of horizontal noise or “banding,” which can obscure the hyperbolic reflections that indicate buried utilities. The raw data will often contain signals from geological layers, soil boundaries, or flat reflectors.

Step 3: Apply the Background Subtraction Filter

In your GPR data processing software, locate the background subtraction filter option. This filter will analyse the data to identify flat-lying reflectors and horizontal bands of noise, and then remove them from the dataset. Here’s how the process generally works:

  1. The software calculates the average response of the data across the horizontal axis (the distance covered in the scan).
  2. The average background noise is subtracted from the entire dataset, leaving only the signals that deviate from the horizontal pattern.
  3. Hyperbolas and other relevant features that represent buried utilities will remain in the data, while the unwanted horizontal reflectors are removed.

Step 4: Fine-Tune the Filter Settings

Adjust the filter settings if necessary. In some cases, the default settings for the background subtraction filter may not fully remove all noise, or they might eliminate valuable data, including useful signals. You may need to adjust the settings manually to fine-tune the filter. Common adjustments include:

  • Window size: This controls how much data is averaged during the subtraction process. A larger window may remove more horizontal noise, but it could also risk removing or losing valuable data.
  • Amplitude threshold: Some software allows you to set a threshold for the amplitude of signals that should be removed. Signals below this threshold are considered noise, while stronger signals (e.g., hyperbolas) are retained.

Step 5: Review the Filtered Data

After applying the background subtraction filter, review the filtered data to ensure that the unwanted horizontal reflectors have been removed. In most cases, you will see a much clearer representation of the subsurface, with the hyperbolas more pronounced and easier to interpret.

Step 6: Interpretation and Utility Mapping

With the horizontal noise removed, you can now focus on interpreting the hyperbolic reflections, which represent the utilities you are trying to locate. Mark the positions of these hyperbolas on your survey, and estimate their depth based on the time it takes for the radar signal to return.

Key Considerations When Using Background Subtraction Filters

1. Avoid Over-Filtering

While background subtraction filters are incredibly useful, be careful not to over-filter your data. Excessive filtering can remove important signals, particularly horizontal utilities or objects that may appear as flat-lying reflectors. Always review the filtered data carefully to ensure that no relevant information has been lost.

2. Understand the Soil and Survey Conditions

The effectiveness of the background subtraction filter can vary depending on soil conditions and the types of utilities you are trying to detect. In environments with highly conductive soils, such as wet clay, you may need to combine background subtraction with other filtering techniques to get clearer results.

3. Combine with Other Filters

In some cases, using a background subtraction filter alone may not be enough to clean up the data completely. Consider combining it with other filters such as band-pass filtering (which removes low and high-frequency noise) or gain adjustments to further enhance the clarity of your GPR data.

4. Caution of Removing Horizontal Utility Data

There are certain cases where you will need to collect and interpret horizontal data and be mindful of eliminating everything from the data set. When a GPR run is completed on top of the length of a utility this can show up as a horizontal band. Data can show the fluctuation as the utility changes depth or holds steady as a linear target, which can be important when planning and designing construction and excavation projects.

Conclusion: Enhancing GPR Surveys with Background Subtraction

The background subtraction filter is an valuable tool in the GPR operator’s arsenal for removing or eliminating unwanted horizontal noise and enhancing the visibility of important features like hyperbolas. 

By eliminating distracting signals from soil layers or boundaries and flat-lying reflectors, this filter allows for clearer detection of utilities, improving the overall accuracy and efficiency of GPR surveys.

When used or applied correctly, the background subtraction filter can significantly reduce the amount of “rubbish data” collected in a GPR survey, allowing you to focus on the key features that matter most—such as buried pipes, cables, and other utilities. 

This ultimately leads to safer excavation practices and more successful projects.

Picture of Simon Williams
Simon Williams

Author of this article

Leave a Reply

Your email address will not be published. Required fields are marked *

Ready to Enhance the Accuracy of Your Utility Detection Projects?

Our expert GPR services can help you achieve cleaner, more reliable data using advanced background subtraction filters. Fill out the form below to schedule a consultation and discover how our GPR solutions can support your project's success!

Discover what our Clients say about us!

Geoscope is proud to be the trusted choice for GPR locating service in Sydney, Australia. But don’t just take our word for it. Here’s what our customers have to say.

Stay Connected

You can also contact us directly via our social media platforms!

Check out our Recent Articles

Watch us on YouTube

You can also connect with us directly via our social media platforms!

Need Assistance?

If you have any questions or need further assistance, feel free to reach out to our team at info@geoscopelocating.com.au.Â