1 Introduction

Impervious surface expansion significantly contributes to environmental degradation. Accurate and efficient mapping of such changes is essential in environmental monitoring and spatial analysis. While many tools exist for land cover classification, few provide a fully open-source, scalable solution capable of ingesting both orthophotos and multispectral data, performing temporal analysis, and supporting integration with spatial databases.
This paper presents a scalable, tile-based system for detecting and monitoring impervious land using orthophotos and Sentinel-2 data. The workflow integrates GRASS for spatial processing, HDF5 for high-performance raster storage, and Python-based machine learning libraries for classification and change detection. Supervised models are trained using labeled vector inputs and applied in a modular pipeline that supports multi-year analysis and large-scale processing. Predicted outputs are postprocessed, vectorized, and stored in spatial databases. Results are further compared with ancillary datasets, including agricultural and forested lands, to inform end users of the environmental implications. The system is designed to support rapid prototyping, reproducibility, and high-resolution monitoring across varying spatial and temporal domains. Performance is evaluated using standard classification metrics, and results demonstrate the applicability of this workflow in operational land change detection contexts.

2 Materials and Methods

2.1 Data Sources

The system supports multiple raster inputs. VHR orthophotos serve as the primary high-resolution data source, complemented by Sentinel-2 multispectral bands. Reference impervious layers from earlier periods generate labeled training data and establish baselines for change detection. Ancillary datasets, such as agricultural and forest land boundaries, are integrated for comprehensive environmental assessments.

2.2 Tiling and Data Management:

Raster and vector layers are preprocessed in GRASS, spatially tiled, and exported to HDF5 format, containing image features and corresponding labels. This enables efficient parallel processing and patch-based classification.

2.3 Classification Pipeline

Machine learning models (Random Forest, GPU-accelerated classifiers, U-Net) are trained using Python libraries such as scikit-learn, cuML, and PyTorch. Training and inference are executed tile-wise using joblib for multiprocessing. Evaluation metrics (F1-score, accuracy) guide model selection and validation.

2.4 Temporal Analysis

Change detection is performed by comparing predicted impervious layers across years. The system computes spatial differences at the tile level to identify newly developed impervious areas. Temporal stacking facilitates scalable monitoring of land surface dynamics.

2.5 Postprocessing and Vectorization

Postclassification filtering reduces noise. Cleaned rasters are vectorized and simplified within GRASS. Final geometries are stored in PostGIS, enabling spatial queries, external integration, and comparison with land-use data for assessing impacts on soil and potential for restoration.

3 Results and Discussion

The system was applied to multi-year orthophoto and Sentinel-2 datasets. Classification accuracy was validated with independent tiles. The HDF5-based tiling significantly reduced I/O bottlenecks. Change maps accurately identified new impervious areas. Vectorized outputs supported further spatial analysis, visualization, and comparative evaluation against agricultural and forested land-use data. These comparisons underscore the importance of monitoring soil sealing and highlight opportunities for targeted land restoration initiatives.

4 Conclusion

This study demonstrates a reproducible, open-source pipeline for high-resolution impervious land mapping. Integration of GRASS, HDF5, and Python libraries enables efficient classification, temporal analysis, spatial output management, and environmental impact assessment. The architecture is adaptable to diverse data sources and modeling approaches, thus suitable for research, operational applications, and informed land management and restoration practices.