Texture-based classification of ground-penetrating radar images
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Abstract Image texture is one of the key features used for the interpretation of radar facies in ground-penetrating radar (GPR) data. Establishing quantitative measures of texture is therefore a critical step in the effective development of advanced techniques for the interpretation of GPR images. This study presents the first effort to evaluate whether different measures of a GPR image capture the features of the data that, when coupled with a neural network classifier, are able to reproduce a human interpretation. The measures compared in this study are instantaneous amplitude and frequency, as well as the variance, covariance, Fourier-Mellin transform, R-transform, and principle components (PCs) determined for a window of radar data. A 50-MHz GPR section collected over the William River delta in Saskatchewan, Canada, is used for the analysis. We found that measures describing the local spatial structure of the GPR image (i.e., covariance, Fourier-Mellin, R-transform, and PCs) were able to reproduce human interpretations with greater than 93% accuracy. In contrast, classifications based on image variance and the instantaneous attributes agreed with the human interpretation less than 68% of the time. Among the textural measures that preserve spatial structure, we found that the best ones are insensitive to within facies variability while emphasizing differences between facies. For the specific case of the William River delta, the Fourier-Mellin transform, which retains information about the spatial correlation of reflections while remaining insensitive to their orientation, outperformed the other measures. Our work in describing radar texture provides an important first step in defining quantitative criteria that can be used to aid in the classification of radar data.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it