Skip to main content

Filter Masks

Filter masks are used to exclude data from the training and testing process. They are useful for removing data that is not relevant to your model, such as saturation masks, cloud masks, etc.

Adding a filter mask​

To add a new filter mask, expand the Filter masks section and click the upload button.

This will bring up an upload dialog where you can drag & drop or select your masks.

Once uploaded, you will see the mask in the Filter masks section.

Filter mask settings​

You can change the following settings for a filter mask:

  • Polarity: If positive, the filter will only include the pixels in the mask, or exclude the pixels in the mask. If negative, the filter will exclude the pixels in the mask, or include the pixels in the mask.
  • Mask type: The type of mask. In datasets, you can select the type of masks to use.
  • Status: If exploratory, the mask will not be used in datasets. If finalized, the mask will be used in datasets. Masks are visible in all workspaces regardless of status.