This update is packed with tools and features. Follow along below to see what's new!
The latest update for Marigold has something for everybody, including a new folder system that allows you to organize your layers more effectively, a tool for querying metadata from the Bare Earth Composites, and improvements to the Vectorize Classified layer tool. We've also streamlined the multispectral and hyperspectral processing tools for a more straightforward user experience. As always, these updates include important changes and fixes to help streamline your Marigold experience.
Folders
Upon launching Marigold 2.7.0, you will immediately notice the introduction of folders within the Raster layers menu. This new feature is designed to assist users in organizing their workspaces, particularly when managing numerous layers. The folder system offers a user-friendly and adaptable approach, allowing you to integrate it into your remote sensing workflow according to your specific needs.
There are multiple ways to add folders to your project:
- Existing project. To bulk organize an existing project with lots of layers, click the
+button next to Raster layers in the left hand panel and selectAdd new folder. In the dialogue that appears on the right, enter a folder name and select the layers you'd like to add, then click Assign. Repeat as needed until all you are happy with your organization.
- Make a new layer. You can individually organize your new layer by clicking the three dot menu next to that layer and select
Folder management. Here you can either assign the layer to an existing folder or type a new folder name to create one.
Folders function similarly to raster layers, providing flexibility in managing your workspace. You can easily rearrange them within the layer order and toggle their visibility on or off using the eye icon adjacent to each folder. Additionally, folders can be collapsed or expanded, helping to minimize visual clutter in your project. To rename or remove a folder, simply utilize the options available in the three-dot menu for each folder. We trust that these features will enhance your organizational experience and contribute to a more streamlined workspace!
Bare Earth Composites (BEC) Query Tool
The Descartes Labs Bare Earth Composites (BEC) from Sentinel-2 and ASTER imagery form the backbone of our mineral exploration products. One common question we receive is about the the original date of each pixel used to create these mosaics. To answer this, we've built a new raster utility to query metadata from the BECs.
To utilize the new utility:
1. Navigate to the Raster management section of the Processing toolbox and select View BEC Metadata.
2. Choose the specific Bare Earth Composite (BEC) you wish to query
3. Click on the Collect metadata button.
4. Then click on the map to place points, which will return the corresponding dates of the Sentinel-2 and/or ASTER scenes used in the BEC.
If you wish to stop querying, simply right-click, move the map, or click the cancel button within the tool.
This new tool can be especially helpful when dealing with areas where there may be large time gaps between the S2 and ASTER, such as when a mine has been developed between the two images.
Consolidated Multispectral and Hyperspectral Tools
A significant enhancement in this release is the consolidation of the classification and transform options within the Processing tools menu. Previously, when Marigold introduced hyperspectral processing capabilities, we created specialized versions of transform and classification tools—such as Principal Components Analysis and Spectral Angle Mapper—designed to optimize the analysis of datasets with hundreds of spectral bands. Importantly, these hyperspectral-specific tools are also compatible with multispectral datasets, including the Bare Earth Composites.
In this release, we have enhanced the HSI versions of these tools, replacing the original multispectral versions. This means that you can now utilize the same MNF tool to effectively process both Fused BEC data and EnMap data. Each of these tools features a streamlined band selection interface, and for some tools, such as Principal Components Analysis, you will also find a detailed statistical breakdown of the analysis results.
Additional changes
- The Vectorize Classified layered tool now identifies classified layers better. This includes results from k-means analyses and the mineral mapping tool.
- Added option to generate screenshots over selected vectors in addition the the previous option of the entire viewport. This helps with exporting irregularly shaped AOIs.
- Fixed issue where geotiff exports did not properly remove regions masked by vectors
- Fixed display issue for vector layers added from the Vector Catalog.