Aller au contenu principal

Raster Processing

In iXGIS, raster data operations are mainly divided into two categories:

  • Raster processing: Performs whole-layer processing on raster layers, such as clip, mask, difference, composite, and mosaic. It usually generates a new result raster layer and leaves the source layer unchanged.
  • Cell editing: Performs detailed editing on cells and requires the Pixel Editor.

On the toolbar, raster processing tools are located on the Raster tab. Cell editing tools are located on the secondary Pixel Editor tab.

To enter the Pixel Editor, select raster data and click Pixel Editor on the Raster tab.

Raster Processing​

astuce

The Clip and Mask operations on this page both use the current map view extent as the spatial boundary. Before running them, pan or zoom the map so the view extent accurately covers the target area.

Clip​

Overview: Clip extracts raster data by the current map view extent, keeps only the part inside the view extent, and outputs it as a new raster layer.

Use cases: Narrowing an analysis area, extracting a region of interest (ROI), and reducing data volume to improve subsequent processing efficiency.

Parameters:

  • Input file: Source raster layer to process. It must already be loaded and selected in the layer list.
  • Storage name: Layer name of the clipped result. Include information such as area or time to make it easier to manage.
  • Storage path: Directory where the clipped result is saved.

Key steps:

  • Select the target raster layer in the layer list.
  • Adjust the map view extent, which is used as the clip boundary.
  • On the Raster tab, click Clip, enter the storage name and storage path, and run the operation.

Notes:

  • The clip boundary comes from the current view extent. For an exact extent, position and zoom the map precisely before running the tool.
  • Clip only generates a new layer and does not modify the source layer, making comparison and rollback easier.

Mask​

Overview: Mask excludes an area from a raster based on the current map view extent. After the operation, cells inside the view extent are usually set to NoData or removed. Raster content outside the view extent is kept and output as a new raster layer.

You can think of it as the opposite of Clip:

  • Clip: Keeps the area inside the view extent.
  • Mask: Removes the area inside the view extent and keeps the area outside it.

Use cases:

  • Quickly removing redundant data within the current view extent.
  • Keeping background or edge areas for overlay or supplemental analysis.
  • Excluding areas from large imagery datasets.
  • Cleaning data and optimizing visualization at the current map view level.

Parameters:

  • Input file: Source raster layer to process. It must already be loaded and selected in the layer list.
  • Storage name: Layer name of the mask result. Include information such as area or time to make it easier to manage.
  • Storage path: Directory where the mask result is saved.

Steps:

  • Open a map project and load data: In iXGIS, open the project that contains the target imagery data. Make sure the relevant imagery layers are loaded in the layer list and display normally in the map view.
  • Adjust the map view: Use zoom and pan tools to position the map window so the view extent accurately covers the area to remove. This view extent is used as the mask area, and cells visible within it will be removed or set to NoData.
  • Select the layer to mask: In the layer list, click the target imagery layer and make sure it is visible. This layer is used as the input data for mask processing.
  • Enter storage path and name: Manually enter the name and save location for the masked raster data.
  • Run the mask operation: Click Mask in raster processing. The system uses the current view extent as the mask area.
  • Generate the mask layer: After processing completes, a new imagery layer is generated and automatically added to the layer list. It displays only the parts not masked by the current view extent.

Notes:

  • The mask boundary comes from the current view extent. Adjust the view accurately before running the operation.
  • The operation applies only to the currently selected layer. Repeat it for multiple layers or use batch processing.
  • Cells in the masked area are usually set to NoData and are excluded from subsequent calculations.
  • Mask only generates a new layer and does not modify the source layer.

Difference​

Overview: Difference performs cell-by-cell subtraction between two raster layers:

Difference raster = minuend raster - subtrahend raster

This operation is commonly used for change detection and time-series comparison, such as NDVI difference and surface elevation change.

Note: The layer selection order directly affects the sign of the result. The layer selected first in the layer list is used as the minuend, and the layer selected second is used as the subtrahend.

Parameters:

  • Input file 1: First raster data used in the difference calculation, as the minuend.
  • Input file 2: Second raster data used in the difference calculation, as the subtrahend.
  • Storage name: Name of the difference result raster.
  • Storage path: Location where the difference result is saved.

Data example:

  • Output raster = target raster - starting raster (Outras.tif = Inras2.tif - Inras1.tif).
  • If either corresponding cell in the two input rasters is NoData, the result cell is NoData.

Steps:

  • Load the two raster layers to compare.
  • Select layers in order: first select the minuend, then select the subtrahend.
  • Enter the storage name and storage path.
  • Click Difference to run the operation. A new difference result layer is generated after completion.

Notes:

  • Difference is a directional operation. Make sure the selection order matches the intended meaning, such as later period minus earlier period.
  • The two input rasters should have consistent spatial reference, resolution, cell alignment, and extent. If not, resampling may be triggered and affect result comparability.

Composite​

Overview: Composite combines multiple single-band raster layers into one multiband raster layer. The result uses the spatial reference and cell alignment of the reference raster, and writes each input layer to a separate band in the selected order.

Use cases: Multispectral compositing, remote sensing preprocessing, and band organization before classification or index calculation.

Parameters:

  • Main file: Raster layer used as the reference. Its spatial reference, resolution, alignment, and extent are used as the basis.
  • Secondary file: Other single-band raster layers included in the composite.
  • Storage name: Name of the output multiband raster after compositing.
  • Storage path: Location where the output result is saved.

Notes:

  • Output band order is determined by the layer selection order. When generating an RGB composite, select layers in red, green, and blue order.
  • Input layers should have the same spatial reference and resolution or cell size, and should maintain cell alignment. Otherwise, resampling may occur.
  • After compositing, check the band order and value range to make sure the result is as expected.

Mosaic​

Overview: Mosaic joins multiple raster layers into one continuous raster coverage and outputs it as a new raster dataset. This process does not modify the source data.

It is widely used in scenarios such as:

  • Joining remote sensing imagery, DEM data, or topographic maps.
  • Building large-area basemaps with cross-region, seamless coverage.
  • Integrating data products collected by tiles or sheets.
  • Visualizing and further analyzing spatially continuous data.

Parameters:

  • Main file: Raster layer used as the reference, usually the first imagery layer loaded or selected.
  • Secondary file: Other raster layers included in the mosaic.
  • Storage name: Name of the mosaic output raster.
  • Storage path: Location where the output result is saved.

Notes:

  • Input layers should have consistent spatial reference, cell size, and alignment. Large differences may trigger resampling and affect image quality and value consistency.
  • Confirm that NoData values are set correctly for each layer to avoid abnormal invalid cell areas after mosaicking.