Raster Reclassify

Reference Palette

The reference palette table on the left side of the dialog shows information about the input layer. This table is option and does not need to be used, but can provide a helpful reference for specify the input and output values in the Raster Reclassification Rules specified on the right table. Use the Load Palette from Layer button to populate the left table with the values present in the input image.

When the left table is populated, one or more rows from the input palette can then be selected and dragged to the populate the Reclass Rules table on the right side. Two options are available for copy values from the left reference palette into the right Raster Reclassification Rules table:

1. Use the Shift or CTRL keys to highlight multiple rows in the reference palette. Drag the highlighted rows to right reclassification table to combine them into a single Output Index value. The default color will be an average of the combined inputs when copying to a blank row.

   Example: The first option is to copy a range of rows from the Input Layer to a single output row in the Reclass Rules. This option is useful to reclassify multiple pixels to a single value. For example, one might want to reclassify all Tree Cover pixels (values 50, 60, 61, 62, 70, 71, 72, 80, 81, 82) to a single pixel value of Trees with a value of 50. Select all the input rows from the reference palette and drag them onto the Reclass Rules. Enter an optional Description in the newly created row. Note that the color of the new row is an average of the colors from the reference palette. If this is not acceptable, you can change the color by clicking on it and choose from the Color Picker dialog.

   

2. Use the Shift or CTRL keys to highlight multiple rows in the reference palette. Drag the highlighted rows to right reclassification table while holding down the CTRL key to move each input layer into the output. By default this rows will maintain the same input Palette Index and Output Index, as well as description and Color.

   Example: The second option is to copy the selected rows from the reference palette into separate rows in the raster reclass rules. To perform a copy, hold the Ctrl key while dragging to rows. See example below.

   

Input Layer – Specify the layer to perform the reclassification on.

Reclassified Layer Name – Specify the name of the resulting layer. By default, this will be the input layer with -Reclassified appended.

Load Palette From Layer– Reads the palette information from the layer and populates the left hand palette list

Load Palette From File – Reads the palette information from a palette file and populates the left hand palette list

Raster Reclassification Rules

The table on the right side of the dialog will be populated with the raster reclassification rules. This table specifies the new output pixel value for each input pixel value or set of input pixel values. This table can be populated by dragging rows from the left table, by right clicking and manually entering new rows, or by using the load button to populate the Raster Reclassification Rules with an existing palette or previously saved Raster Reclassification Definition.

• Palette Indices – Specify the input pixel values that will be converted to the output. This can be listed as a single value, comma separated values, or a range of values using a - . It is also acceptable to add a row containing the keywords else, other or default (case insensitive) to capture all input pixel values not specified in another row.

• Output Index – Specify the output Palette Index to be applied when reclassifying the input Palette Indices.

• Description - Optionally specially a palette description.

• Color - Specify a color for the palette value in the output image. Click the color to bring up the Color picker dialog, which also includes a choice to Pick Color from Map.

  

The reclass rules can be edited manually to create reclassification rules. Double click on a Reclass Rule cell to edit it. Use the Return key to complete editing, or the tab and arrow keys to move to different cells. The Insert key add a new blank row to the reclass rules grid.

In addition, right-clicking the mouse in the Reclass Rules grid will bring up a menu option to add or delete a row. As shown in image below, the context menu is primed to delete row with Palette Indices 70-72, or to insert a new row at the bottom of the list.

When editing the reclass rules, the following conditions must be followed.

Condition	Example
A Palette Indices row may contain valid palette index values separated by a comma.	1,5,7,10
A Palette Indices row may specify a range contain valid palette index values separated by a dash.	70 - 72
A Palette Indices row must not contain both individual values and ranges	1,5,7,10,70 - 72 (Invalid)
Reclass Rules may contain a single keyword for the default case indicating that if no other rule applies then use the Output Index value for a pixel conversion.	Acceptable keywords: ELSE OTHER DEFAULT
A Palette Index may only appear in one row.	1,5,7,10,11,12,13 12-20 (Invalid because the range 12-20 contains values 12 and 13)
An Output Index must contain a single valid Palette index or one of the keywords for No Data. The No Data value will be assigned the current transparent color for the layer.	50 Acceptable keywords: NODATA NO DATA TRANSPARENT ND

Load – This loads in the palette or saved Raster Reclassification Definition file specified in the drop-down to the right and populates the Raster Reclassification Rules table below.

Save – Saves the Raster Reclassification Rules specified in the table to a Raster Reclassification Definition file that can be reused for future reclassifications. This is a *.gmr file that is a comma separated list of the values specified in the table that list the input Palette Indices, Output Index, Description and RGB Color value.

Validate – Checks the Raster Reclassification Rules and combines values into ranges where appropriate. To ensure the rules are correct, press the Validate button. This will check that all the conditions in the above table are met, and group individual Palette Indices together for equal Output Index values. Validating will not only check the data, but also sort it by Output Index.