The process for creating mosaic datasets will vary depending on the complexity of your image dataset.
If you have a single data collection (e.g. data from 2010), you'll create a single source mosaic dataset, which you can then share directly, or create a cache to share.
If your data has multiple collections (e.g. data from 2005, 2010, and 2015), you'll create source mosaic datasets for each data collection, then combine them into one derived mosaic dataset that can then be shared.
The Create Source Mosaics tool automates many of the best practices outlined here. These specifications are provided mainly for reference.
Source mosaic datasets
How should I set up my mosaic dataset?
The Create Source Mosaic Dataset workflow tool automatically sets the appropriate mosaic dataset properties, or prompts the user to do so. The recommended settings are listed here (any properties not mentioned should use the default setting). To set these properties manually, run the Set Mosaic Dataset Properties tool.
(Transmission Compression setting) Default Method | JPEG |
(Transmission Compression setting) JPEG with Recommended Quality | 80 |
Default Resampling Method | Bilinear interpolation |
Download | To allow users to download source raster data, set this to maximum number of downloadable rasters to limit the number of images. To disable download, set this value to 0. |
Blend Width | 0 |
Always Clip the Raster to its Footprint | Yes |
Footprints May Contain NoData | No |
Always Clip the Mosaic Dataset to its Boundary | No |
Statistics | Not required |
(Raster Information) Source Type | Processed (i.e. color correction and contrast enhancements have been applied, so an automatic stretch will not be applied.) |
Note:
If you run Calculate Values to set properties, review the overview records in the attribute table to ensure they include the same key metadata as the rest of the image collection.
What coordinate system should I use for my mosaic dataset?
If the source data uses one projection | Set the coordinate system for each source mosaic dataset to be the same as the source data. This will simplify quality assurance/quality control testing. |
If a collection includes multiple projections (for example, in states that cross multiple UTM or state plane zones) | Either select the zone that covers the most datasets within the collection, or use a projection better suited for all the data, such as Web Mercator. |
- The mosaic dataset performs projection on the fly, so the projection in which the mosaic dataset is managed does not need to be the same as that of the original data.
- Reprojection will be performed as a single step, from original source data to desired output projection, minimizing any resampling.
What parameters should I use when adding rasters?
See documentation on adding rasters to mosaic datasets. Use the following parameters when adding preprocessed orthophotos.
| Parameter | Recommended Settings |
|---|---|
Cell Size Tolerance Factor (in mosaic dataset Properties) | 999 |
Raster type | Raster Dataset |
Update Cell Size Ranges flag | Off |
Update Overviews flag | Off |
- Records in the source mosaic dataset can be selected and merged after they've been added, combining them into a single record in the attribute table. This may be simpler for end users who will have access to multiple image collections.
- It is better to add overviews later, after key metadata has been added.
How should I populate metadata?
First, add appropriate fields to the mosaic dataset attribute table. Recommended names and data types for aerial imagery metadata are listed here. Record these attributes for each source mosaic dataset-the metadata will then be copied into any derived datasets. Users can also add other metadata fields to mosaic datasets as needed.
| Field Name | Data Type | Description |
|---|---|---|
Year | Long integer | This field is commonly used for defining which image in a mosaic dataset appears on top |
FirstDate | Date | Record the earliest acquisition date for the raster dataset (if unknown, use January 1, YYYY). Note:Raster dataset may contain data from more than one image, and the acquisition date for individual images may not be available. Instead, it is helpful to have the date range of acquisition dates for all the contained images. |
LastDate | Date | Record the latest acquisition date for the raster dataset (if unknown, use December 31, YYYY). (See FirstDate) |
Dataset_ID | Text16 | A unique ID for each data collection |
Source | Text24 | The original source of the data (e.g. USGS, or the data provider for a custom project) |
Metadata_URL | Text64 | Direct users to the location of original metadata files stored on an accessible network location |
SensorName | Text20 | The sensor used to capture the collection (e.g. the digital camera name or film camera type) |
Leaf_Status | Short | Identify leaf-on or leaf-off imagery (1=Leaf_On, 2=Leaf_Off, 0=Unknown) |
Second, populate the fields. If all the data in a collection have the same metadata, use the Calculate_Field tool to fill in the values.
How should I deal with NoData values in my imagery?
NoData pixels on the edges of orthophotos can make it difficult to mosaic your imagery. The Create Source Mosaics workflow tool is designed to streamline the process of removing NoData pixels using the Footprintparameter. Unless your tiles are edge-matched, set the Footprint parameter to "Remove black edges around imagery" to remove NoData pixels.
To remove NoData values manually, you will need to do the following.
- 1. Set the mosaic dataset properties as follows:
Clip to Footprint?
Yes
Define NoData value?
No
- Clip your orthophotos to a footprint (i.e. accurate polygons that define the valid extents of imagery within each file, like a polygon representing the county boundary). You can either (A) use existing footprints or (B) build them with the Build Footprints tool. Follow the guidelines below for your type of data.
If you have accurate footprints: Import those polygons into the footprint of the mosaic dataset using the Import Footprint or Boundary tool.
If you do not have footprints: Run the Build Footprints tool using the Radiometry option and the following parameters, depending on the format of your data.
Orthophoto format Number of vertices Shrink distance Tiled orthophotos
8
0
Multi-image mosaics
Reasonable estimate depending on footprint complexity (~100)
Large enough to remove most, but not all, of the overlap between images
Individual, orthorectified image frames
4
Set to a value that will reduce the footprint to be well inside the valid extents of each image, but without losing the overlap between images. If the source data files have a collar of non-image data (such as fiducials and film annotation, if the source was scanned film), you will likely need to use a much larger shrink distance; 200 pixels, for example.
Note:
The units for shrink distance are based on the coordinate system of the mosaic dataset, not pixels, so consider the resolution of your imagery (meters per pixel) when setting this. For example, with 0.5-meter data, a value of 50 meters is equivalent to 100 pixels.
If the data is compressed (e.g. MrSID, ECW, JP2, or JPEG formats), the original NoData values may have been altered. As a result, the Build Footprints process must be configured to ignore a range of NoData values. If the NoData is black (pixel = 0), adjust the minimum value to a few digits greater than zero (for example, try 4). If the NoData is white (pixel = 255), adjust the maximum value to a few digits less than 255 (for example, try 251).
You will probably need to experiment to determine the optimum settings for your data. If you have a large collection of imagery, test with a small selection of images before processing your full dataset.
Do I need to apply color correction?
Color correction has likely already been applied to preprocessed orthophotos, in which case additional color correction (refining image radiometry) isn't necessary.
If color correction has not been applied, examine the imagery visually to decide if it is important. If you do need to apply color correction, correct each source mosaic dataset separately (for image-to-image matching). You may then decide to adjust the source mosaic datasets for year-to-year matching. For more information about how to apply color corrections, refer to Color Balancing a Mosaic Dataset.
Note:
Color correction can't be applied to multi-image mosaics. If the data shows color or contrast problems, this will need to be fixed in the individual images by your data provider before building the multi-image mosaics.
Do I need to create seamlines?
Seamlines are not required unless you are working with individual orthorectified image frames. Examine your source mosaic visually, and if you find distracting mismatches between images (typically from tall buildings), you can build seamlines.
Do I need to create overviews?
Overviews aren't necessary if you are creating a raster tile cache with your imagery.
Otherwise, you should create overviews for each source mosaic dataset.
- Run the Define Overviews tool using the following parameters.
Parameter Value Base pixel size
Approximately 1/500 of the width of a typical image file. For example, if an image covers two kilometers, set the overview base pixel size to four meters.
Compression
JPEG
Compression quality
80
Resampling
Bilinear
- After overviews are defined, run the Build Overviews tool.
Derived mosaic datasets
If you are working with multiple data collections (i.e. more than one mosaic dataset), you may want to merge your source mosaic datasets into a single derived mosaic dataset to streamline data management. With all data collections managed in a single derived mosaic dataset, user applications can easily search the attribute table to access image metadata, and then select imagery by date or any other attribute.
How should I set up my mosaic dataset?
The Create Derived Mosaic Dataset workflow tool automatically sets the appropriate mosaic dataset properties, or prompts the user to do so. The recommended settings are listed here (any properties not mentioned should use the default setting). To set these properties manually, run the Set Mosaic Dataset Properties tool.
When creating your derived mosaic dataset, define it according to the following guidelines:
- Define the number of bands and pixel depth for the derived mosaic dataset based on the maximum number of spectral bands, and the maximum pixel depth, of any source dataset.
- Set the spatial reference for the derived mosaic dataset to accommodate the maximum extents expected for your project. Web Mercator Auxiliary Sphere is typically used.
For the derived mosaic dataset, the recommended properties settings are:
| Property | Setting |
|---|---|
(Transmission compression setting) Default method | JPEG |
(Transmission compression setting) Recommended quality | 80 |
Default resampling method | Bilinear Interpolation |
Allowed mosaic methods | Lock Raster, By Attribute, Seamline, and None |
Default mosaic method | By Attribute |
Order Field | Year |
Order | Decreasing |
Base Year | 3000 |
Maximum size of requests | The recommended maximum is 4000 x 4000 to restrict users from capturing a local copy of your full data collection. Default is 15000 x 4100. |
Download | If users require source data, and your network bandwidth is adequate, set Maximum Number of Items Downloadable to 20. Otherwise, set Maximum Number of Items Downloadable to 0 to disable download. |
Blend Width | 0 |
Always Clip the Raster to its Footprint | Yes |
Footprints May Contain NoData | No |
Always Clip the Mosaic Dataset to its Boundary | No |
Statistics | Not required. Nominal values for the statistics covering the entire project area should be stored in a configuration file and imported into the derived mosaic dataset using the Set Raster Properties function. |
(Raster Information) Source Type | Processed |
Note:
These settings are intended for users managing collections of imagery from different years who want to display the most recent image.
How do I add my source mosaic datasets to my derived mosaic dataset?
All data added to the derived mosaic dataset should be contained within source mosaic datasets using the parameters below (no individual rasters should be added to the derived mosaic dataset).
| Parameter | Setting |
|---|---|
Raster type | TABLE |
Calculate Cell Sizes | Off |
Update Boundary | On |
How should I populate metadata?
After adding the source mosaics, ensure Dataset_ID and other key metadata fields have been copied into the derived mosaic dataset for all records.
How should I deal with footprints and NoData values in my imagery?
The source mosaic datasets should have footprints already defined. Derived mosaic dataset properties should be set as follows:
| Property | Setting |
|---|---|
Clip the raster to its footprint | Yes |
NoData value | Do not define |
Do I need to create overviews?
Yes. Set the base pixel size should to a value approximately equal to the average extent of a full data collection divided by 500. For example, if you are managing imagery from multiple states (each in its own source mosaic dataset), and Kansas is included (~650 km east to west), use 1300 meters to define the base pixel size for overviews.
Note:
Small-scale views generated from high-resolution imagery can appear bland. As an alternative, you can use an independent data source better scaled to provide geographic context-for example, a mosaic of Landsat data. If you go this route, create an additional source mosaic dataset for the lower-resolution imagery (with appropriate overviews), and then add that to the derived mosaic dataset to provide the small-scale views.
Creating on-the-fly products
Users often want several different representations of image data. With preprocessed orthophotos, for example, users might want RGB color, color infrared, and Normalized Difference Vegetation Index (NDVI), each selectable by year. Rather than writing more data to disk, it's often better to create alternative products using on-the-fly processing functions. See here for more information about the available raster functions and how to use them.
Note:
This is an option for datasets with more than three bands, or multi-year collections. If you will be serving cached tile imagery, this is generally not applicable.
Check out the next section to learn more about best practices for publishing in the Managing Preprocessed Orthophotos workflow.