Added Section Preview for Attribute Selection  

The new section preview feature simplifies selecting the attribute best suited for the GPR survey purpose. Previously, users would create full-size sections to test different attributes, which could be time-consuming. Now, users can quickly select the desired section in the preview window using a priori knowledge or subsurface layer patterns. The attribute of the selected preview section is automatically applied for the final section calculation. Preview generation time depends on the GPR profile size, but it is significantly shorter than creating full-size sections for attribute selection. 

In addition to section images, the preview window includes controls for adjusting and saving section color schemes, an auto-contrast feature to enhance low-contrast sections, and a screenshot option to save the preview window image.  

Added Section Preview for Corrective Function Selection

The preview feature for the primary attribute, the Real part of complex relative permittivity, streamlines corrective function selection, which is used to reduce errors in automated backscatter field analysis (BSEF). Preview sections are generated using various types of corrective functions. Users can select the most suitable section for their survey in the preview window. 

The corrective function used for this section is then applied to the final section calculation. The corrective function preview window offers the same color scheme adjustment and other settings as the attribute selection preview window.

New Data Cleaning Method Added 

A new BSEF analysis data cleaning method using the interquartile range (IQR) method has been added. Data cleaning using the IQR method identifies attribute values that fall outside the range [Q1 - K*IQR, Q3 + K*IQR], where Q1 is the first quartile, Q3 is the third quartile, IQR (Interquartile Range) is calculated as IQR=Q3−Q1, and K is the IQR multiplier. Analysis points with attribute values outside this range are considered outliers and are excluded from the dataset used for section calculation. The user, by varying the values of the IQR multiplier K, can adjust the width of the interquartile range, thereby controlling the sensitivity of the method to outliers of the attribute values. Parameters for data cleaning can be configured in the IQR-Based Data Cleaning panel, located in the left tab group.

Optimization Option Renamed 

The Optimization option has been renamed to Mean Deviation Data Cleaning to better reflect its functionality. Accordingly, the Optimization panel is now ​titled Mean Deviation Data Cleaning. Other parameter names within this function remain unchanged. 

New Data Cleaning and Attribute Limit Reset Buttons Added 

Separate buttons for resetting data cleaning actions and base attribute limits are now available on the new Cancel Actions panel in the left tab group. The panel includes the following buttons:
- Range limit cancels all base attribute range limits. This duplicates the Max Rng All button on the Permittivity histogram panel.
- Data cleaning undoes data cleaning results, useful for reverting to the original BSEF analysis results after applying multiple cleaning methods. Instead of using separate reset buttons for each method, this button allows all changes to be undone at once.
- Cancel all resets all restrictions with one click, including data cleaning results and base attribute ranges.

Added Option to Select Base Attribute Smoothing State 

BSEF automated analysis results now include base attribute data in both original and smoothed states. Base attribute smoothing in the analysis process minimizes potential analysis errors caused by sharp anomalies in attribute values. Previously, unsmoothed data was not retained in the BSEF analysis results. Now, to select the data smoothing state, users can use the Smooth checkbox on the Permittivity histogram panel.

Unsmoothing the analysis results provides a more detailed subsurface model, useful for tasks requiring fine-grained object detection or localized subsurface changes. However, this level of detail may complicate interpretation, as minor fluctuations can create visual "noise" and distractions.  

For tasks focusing on the overall subsurface structure, the smoothed version of the analysis results is preferable. Smoothing minimizes local value fluctuations, making the layers in the section appear more continuous. This helps highlight layer boundaries and makes it easier to analyze subsurface structures, especially in highly heterogeneous environments or when the data is noisy.