MassGIS 2023 Aerial Imagery Dataset

Dataset Description

This dataset contains high-resolution "leaf off" aerial photographs covering the entire state of Massachusetts, captured in spring 2023. The imagery has been converted from specialized formats into standard formats suitable for AI and machine learning applications.

Overview

The MassGIS 2023 Aerial Imagery Dataset provides comprehensive aerial photography coverage of Massachusetts, captured during leaf-off conditions to maximize ground visibility. Originally acquired as 10,218 individual tiles in GeoTIFF format and later compressed to JPEG 2000, this dataset has been processed to create standardized, machine-learning-ready imagery that supports a wide range of geospatial AI applications.

Source Data

• Original Source: MassGIS Data: 2023 Aerial Imagery
• Original Format: JPEG 2000 files (converted from GeoTIFF) in a U.S. National Grid (USNG) tile structure
• Capture Date: Spring 2023 (February 21 to April 28)
• Resolution: 15 cm (6-inch) ground resolution
• Bands: 4-band (RGB-IR) imagery
• Spatial Reference: NAD 1983 (2011) UTM Zones 18N and 19N (meters)
• Accuracy: ASPRS Class 1 horizontal accuracy (tested 0.26-0.385 meters RMSEr)
• Data Owner: MassGIS (Bureau of Geographic Information)

Acquisition Details

The aerial imagery was acquired using an UltraCam Eagle M3 camera with a flight design that included a total of 214 flight lines to capture 11,158 photos. Imagery was collected on 15 days between February 21 and April 28, 2023, under conditions free from clouds, shadows, smoke, fog, haze, snow, ice on water bodies, flooding, and excessive soil moisture.

Funding for this imagery was provided by MassDOT and the State 911 Department. The imagery was originally delivered as 10,218 individual 1,500 m × 1,500 m tiles in GeoTIFF format (763 MB per tile) and later compressed by MassGIS into JPEG 2000 format (approximately 19 MB per tile) for easier distribution.

Original Production Process

The orthoimagery creation process included:

1. Aerial Acquisition: Imagery was captured with an UltraCam Eagle M3 camera and supplemented with airborne GPS/IMU data.

2. Raw Image Processing: Raw images were radiometrically and geometrically corrected using UltraMap software, then manually edited to ensure appropriate tone and seamless blending between images.

3. Orthophoto Creation: Using Hexagon Geomedia, images were draped upon a ground model derived from various LiDAR bare earth datasets and orthorectified. Manual seamlines were drawn in ArcMap on every frame.

4. Accuracy Assessment: Horizontal accuracy was tested using 64 independent ground control points. The results demonstrated ASPRS Class 1 accuracy with RMSEr values of 0.385 meters for UTM Zone 18 and 0.26 meters for UTM Zone 19.

5. Compression: MassGIS converted each GeoTIFF to an 8-bit JPEG 2000 (GMLJP2) format using GeoExpress 10 software at a 20:1 lossy compression ratio, maintaining all four bands while reducing file size from 763 MB to approximately 19 MB per tile.

Dataset Structure

The dataset is organized by geographic zones and includes:

• Full-resolution imagery: High-quality images suitable for detailed analysis, converted to JPG format
• Preview imagery: Downsampled versions for quick exploration
• Metadata: Information about each image tile including:
  • tilename: Unique identifier for each tile based on USNG coordinates
  • zone: Geographic zone identifier (UTM Zone 18N or 19N)

Band Information

The original imagery contains 4 bands that can be rendered in different combinations:

• Natural Color: Bands 1, 2, 3 (RGB) assigned to red, green, blue channels
• Color Infrared: Bands 4 (near infrared), 1, 2 assigned to red, green, blue channels
  • In color infrared view, vegetation appears red
  • Artificial surfaces (like artificial turf) that appear green in natural color view take on a grayish hue in infrared

Intended Uses

This dataset is designed for:

• Environmental monitoring and land use change detection
• Urban planning and infrastructure management
• Forestry and agricultural analysis
• Machine learning model training for object detection in aerial imagery
• Geospatial AI applications
• Emergency management and disaster response planning
• Building footprint extraction
• Land cover classification

The imagery was specifically created "to provide easily accessible geospatial data which are readily available to enhance the capability of federal, state, and local emergency responders, as well as to plan for homeland security efforts."

Limitations

• While the imagery was captured in leaf-off conditions, some deciduous vegetation may still obstruct ground features
• Slight systematic radiometric differences may exist between adjacent orthoimage files due to differences in source image capture dates and sun angles
• Seamline mismatches may be present if the displacement was less than one pixel
• The imagery represents a specific time period (Spring 2023)
• Urban areas with tall buildings may have perspective distortions
• Water features may have sun glint in some images
• Some areas may have been affected by seasonal conditions

Citation

If you use this dataset in your research or applications, please cite:

Massachusetts Bureau of Geographic Information (MassGIS). (2023). 
MassGIS Data: 2023 Aerial Imagery [Data set]. 
Retrieved from https://www.mass.gov/info-details/massgis-data-2023-aerial-imagery

License

This dataset is made available under the same terms as the original data published by MassGIS, which has no restrictions. As noted in the original documentation: "No restrictions apply to these data. However, users should be aware that temporal changes may have occurred since this dataset was collected and that some parts of these data may no longer represent actual surface conditions."

Acknowledgement of MassGIS would be appreciated for products derived from these data.

Acknowledgements

We thank the Massachusetts Bureau of Geographic Information (MassGIS) for making the original aerial imagery publicly available. The original imagery acquisition was funded by MassDOT and the State 911 Department, with planning, acquisition, processing, and creation of derivative products by MassGIS and NV5 Geospatial of Lexington, KY.