PAR Government Systems Corporation (PGSC) excels in ground classification of remotely-sensed imagery using multi-resolution and multi-spectral imagery. A combination of automated techniques and advanced Geographic Information System (GIS) methodologies allows PGSC to meet customer requirements for specific applications such as land cover/land use classification, feature extraction, water quality modeling, and flood mapping.
Application of multi-resolution, multi-source imagery to classification is a cutting-edge technique providing scalable products to meet changing demands over a lifetime of customer needs. Detailed classifications may be used for complex modeling tasks or scaled back to meet generalized mapping requirements.
These images display a portion of the automated classification developed for the New York City Department of Environmental Protection (NYC DEP) on-going efforts within the New York drinking watershed. The base classification is generated from LandSAT 7 ETM+ imagery.
Additional land use products are developed using a combination of the 25-meter ETM+ imagery and EmergeTM 0.3-meter imagery in conjunction with ancillary datasets such as NWI wetland digital files and town tax parcels.
Methods
- Supervised Classification
- Texture Analysis
- Frequency Analysis
- Resolution Merging
- Sub-pixel Classification
- Feature Extraction
Software Tools
- ArcGIS
- ERDAS IMAGINE
- Feature Analyst
- PGSC Custom Tools
LandSAT ETM+ 25 meter natural color image subset
Emerge™ 0.3 meter color-infared image subset
