Automated Building Information Extraction and Evaluation from High-resolution Remotely Sensed Data
Bibliographic record
Abstract
The two-dimensional (2D) footprints and three-dimensional (3D) structures of buildings are of great importance to city planning, natural disaster management, and virtual environmental simulation. As traditional manual methodologies for collecting 2D and 3D building information are often both time consuming and costly, automated methods are required for efficient large area mapping. It is challenging to extract building information from remotely sensed data, considering the complex nature of urban environments and their associated intricate building structures.\nMost 2D evaluation methods are focused on classification accuracy, while other dimensions of extraction accuracy are ignored. To assess 2D building extraction methods, a multi-criteria evaluation system has been designed. The proposed system consists of matched rate, shape similarity, and positional accuracy. Experimentation with four methods demonstrates that the proposed multi-criteria system is more comprehensive and effective, in comparison with traditional accuracy assessment metrics.\nBuilding height is critical for building 3D structure extraction. As data sources for height estimation, digital surface models (DSMs) that are derived from stereo images using existing software typically provide low accuracy results in terms of rooftop elevations. Therefore, a new image matching method is proposed by adding building footprint maps as constraints. Validation demonstrates that the proposed matching method can estimate building rooftop elevation with one third of the error encountered when using current commercial software.\nWith an ideal input DSM, building height can be estimated by the elevation contrast inside and outside a building footprint. However, occlusions and shadows cause indistinct building edges in the DSMs generated from stereo images. Therefore, a “building-ground elevation difference model” (EDM) has been designed, which describes the trend of the elevation difference between a building and its neighbours, in order to find elevation values at bare ground. Experiments using this novel approach report that estimated building height with 1.5m residual, which out-performs conventional filtering methods.\nFinally, 3D buildings are digitally reconstructed and evaluated. Current 3D evaluation methods did not present the difference between 2D and 3D evaluation methods well; traditionally, wall accuracy is ignored. To address these problems, this thesis designs an evaluation system with three components: volume, surface, and point. As such, the resultant multi-criteria system provides an improved evaluation method for building reconstruction.
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 | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| 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".