Illuminated Road Extraction Based on Deep Learning Using the Nighttime Light and Multispectral Imageries of SDGSAT-1
Bibliographic record
Abstract
The 10-m nighttime light (NTL) imagery provided by the Sustainable Development Science Satellite-1 (SDGSAT-1) provided an effective data source for the extraction of SDGSAT-1 illuminated roads on the Earth. Although our previous work introduced an improved U-Net model for illuminated road extraction from the NTL imagery of SDGSAT-1, some of the illuminated roads extracted from the NTL imagery were disconnected due to relatively poor illumination conditions. To improve the accuracy of extracting illuminated roads using NTL imagery, this study used both NTL imagery and daytime multispectral (MS) imagery taken by SDGSAT-1. Several state-of-the-art deep learning models were considered in this work, including U-Net, SegNet, ResUNet, D-LinkNet, DeeplabV3+, and DSC-ResUNet. The predicted results of these models using the combination of NTL and MS (NTL+MS) as the input were compared with those obtained using only the NTL bands as the input. The Precision, Recall, F1, and IoU values of Deeplab V3+, D-LinkNet, Unet, ResUNet, and DSC-ResUNet obtained by using the combination of NTL and MS as the input were higher than those obtained by using only the NTL bands as the input. Meanwhile, the visual comparison also showed that the predicted results of NTL+MS were more accurate and reduced the cases of missing and disconnected road segments. The experimental results demonstrated that the combination of NTL and MS images is effective for improving the accuracy of illuminated road extraction using deep learning.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.002 |
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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".