Mosaicking
The process of combining multiple satellite images into a single unified composite image covering a larger area. Involves color balancing, seam line removal, and blending to create a visually consistent output.
Overview
Mosaicking combines multiple overlapping satellite or aerial images into a single unified composite covering a larger area. Because satellite sensors capture data in discrete tiles, any analysis spanning a region requires merging multiple acquisitions. The challenge is that adjacent scenes may be acquired on different dates under different conditions, producing visible seams and color discontinuities.
How It Works
Creating a mosaic involves four steps: image selection (choosing best available pixels based on cloud cover and quality), geometric alignment (ensuring consistent coordinate reference), color balancing (adjusting radiometry of overlapping images to minimize brightness differences), and seam handling (placing seams along natural features and using feathering to smooth transitions).
For cloudless composites, algorithms analyze image stacks over time windows and select the best pixel at each location. Google Earth Engine's median composite and Sentinel Hub's least-cloudy-pixel approaches are common implementations.
Key Facts
- Pixel-based compositing has largely replaced traditional scene-based mosaicking for analytical applications.
- Color balancing must preserve spectral fidelity for downstream quantitative analysis.
- Seam lines through homogeneous areas (forests, water) are less visible than through high-contrast boundaries.
- Google Earth Engine, Sentinel Hub, and Planetary Computer provide on-the-fly compositing capabilities.
- Seasonal mosaics require careful date selection to avoid mixing phenological stages.
Applications
National and Global Basemaps
Google Maps, Bing Maps, and ESRI World Imagery assemble global mosaics from satellite imagery.
Wall-to-Wall Land Cover Mapping
National land cover programs require continuous coverage without gaps between tiles.
Cloud-Free Composite Generation
In persistently cloudy regions, pixel-based compositing from time stacks achieves complete coverage.
Agricultural Monitoring at Scale
Continuous regional imagery ensures uniform crop condition tracking across administrative boundaries.
Limitations & Considerations
Temporal compositing produces images representing no single moment in time, creating inconsistencies in vegetation state or shadow direction. Color balancing can alter absolute reflectance values. Persistent cloud cover may leave residual gaps. Feathering can create blurred artifacts where land cover changed between acquisition dates.
History & Background
Photomosaics have been produced since the 1910s-1920s. Digital mosaicking emerged in the 1970s with satellite imagery. NASA's GeoCover produced one of the first global Landsat mosaics. Google Earth (2005) popularized continuous global mosaics. Modern approaches increasingly use median pixel selection from dense time series.
Analyze Mosaicking data with LYRASENSE
Use our agentic notebook environment to work with satellite data and compute indices like Mosaicking — no setup required.