Multispectral Data Processing
      Back to home
      1. Wingtra Knowledge Base
      2. 4. Process Your Data
      3. Multispectral Data Processing

      Process RedEdge Panchromatic images with Metashape

      Learn how to process MicaSense RedEdge Panchromatic (RE-P) images with Agisoft Metashape photogrammetry software.

      General recommendations

      • WingtraOne datasets, in particular those that are captured with the MicaSense RE-P sensor, can be quite large. Please make sure to have enough computing resources available.
      • We recommend >= 32GB RAM, >6 core and enough high-speed SSD storage.

      💡 At the moment, Metashape is not directly compatible with PPK-geotagged images. If you ran PPK geotagging, we recommend using the raw images collected by the drone and importing the geotags CSV file for more precise location information.

      Step 1. Create the project and import the images

      • Open Metashape and create a new project (File -> New)
      • Open the Workflow menu and select Add Photos or Add Folder to import all the images
      • Select Multi-camera system
        Multi-camera system highlighted

      Metashape identifies the calibration images and moves them to a special camera folder so that they are not used in the alignment. 

      Step 2. Set primary channel to Panchro band

      To use the higher resolution panchromatic band to align the data, make sure that you set the primary channel to Panchro under the Tools -> Set Primary Channel menu.

      Set primary channel 1

      In the Set Primary Channel window select Channel 3 - Panchro and then click on Apply and OK.

      Panchro primary channel

      Step 3. Calibrate camera

      In order to compare the multispectral data to other measuring tools, to satellite imagery, or even across different times of the season, it is important that the images are calibrated radiometrically. The images of the calibration panel that were taken before and after the flight are used for this step.

      To implement the calibration, go to the Tools menu and then select Calibrate Reflectance...

      Calibrate reflectance_highlighted

      In the window that pops up, select Locate Panels, and the software will look for the pictures of the reflectance panel. It will also show the corresponding reflectance values for each band.

      Calibrate reflectance_2_new

      Note that a warning message indicating a missing panel calibration value will come up. This is because the panchromatic band does not need reflectance calibration.

      Make sure you check the Use reflectance panel options at the bottom of the window. For sun sensor calibration, we only recommend this if the drone was flown under overcast conditions. If the skies were clear, reflectance panel calibration is sufficient.

      Set Brightness: Go to Tools in the menu and select Set Brightness.

      Image brightness

      In the Image Brightness window, select Estimate and then Apply.

      Step 4. Align photos

      Open the Align photos ... dialog from the Workflow menu.

      align photos_highlighted

      We recommend using High accuracy and that you tick the Generic preselection and Reference preselection boxes.

      align photos options

      Once this step is completed, in the Model view you can see the computed camera locations and tie point cloud.

      Step 5. Build dense cloud

      Open the Build Dense Cloud dialog from the Workflow menu.

      dense point cloud_highlighted

      We recommend using High Quality for better results. Note that this can increase the processing time.

      Step 6. Build DEM

      Open the Build DEM dialog from the Workflow menu.

      build DEM

      We recommend using High Quality for better results. Note that this can increase the processing time.

      Step 7. Build orthomosaic

      Open the Build Orthomosaic dialog from the Workflow menu.

      Build orthomosaic_highlighted

      Use DEM as a source surface. Note that the software will take some time since it needs to create 6 orthomosaics (one for each band).

      You can view the Orthomosaic by double-clicking on Orthomosaics under Chunk1 

      Step 8. Pansharpening

      Pansharpening can be done in Metashape using the raster calculator. Pansharpening uses a higher-resolution panchromatic image fused with lower-resolution multiband images to create a higher-resolution raster dataset.

      Open the Raster Calculator from the Tools menu and use the formulas below to pansharpen the bands for a pansharpened version of the dataset.

      raster calculator

      You can copy the formulas below:

      B1 * (B3 / (0.2 * B1 + 0.2 * B2 + 0.2 * B4 + 0.2 * B5 + 0.2 * B6)) / 32768

      B2 * (B3 / (0.2 * B1 + 0.2 * B2 + 0.2 * B4 + 0.2 * B5 + 0.2 * B6)) / 32768

      B4 * (B3 / (0.2 * B1 + 0.2 * B2 + 0.2 * B4 + 0.2 * B5 + 0.2 * B6)) / 32768

      B5 * (B3 / (0.2 * B1 + 0.2 * B2 + 0.2 * B4 + 0.2 * B5 + 0.2 * B6)) / 32768

      B6 * (B3 / (0.2 * B1 + 0.2 * B2 + 0.2 * B4 + 0.2 * B5 + 0.2 * B6)) / 32768

      Step 9. Export the orthomosaic

      To export the orthomosaic with the above raster transformation, select on the File menu the Export Orthomosaic -> Export JPEG/TIFF/PNG option. In the Export Orthomosaic page, set the Raster transform field to Index Value and Export the orthomosaic.

      export orthomosaic_highlight

      If you don't want to create a pan-sharpened dataset, you can choose "None" in the Raster transform field.

      If you are interested in creating additional multispectral products from RE-P data, you can find more information in the article Creating multispectral products with RedEdge-P and QGIS.