Export Confusion Matrix

Export a confusion matrix based on the comparison of an image object classification to defined sample objects or object variable.

Related to algorithm Compute Accuracy Values and Tools > Accuracy Assessment to calculate and visualize a confusion matrix. Description of workflow in User Guide Developer > Tools > Tools.

See also this video - Accuracy Assessment

Related to this tutorial - Sample Statistics and Accuracy Assessment Tool

Supported Domains

Execute;

Algorithm Parameters

Export Mode

• Static export item writes the export item to the workspace. A rollback will remove the exported information. Enter a name in Export Item Name, or use the default name
• Dynamic export item lets you enter or select a variable as an export item. Select a variable from the drop-down box or enter a name in the Export Item Variable Name field (entering a name launches the Create Variable dialog box, where you can enter a value and variable type) . The export is written to the scene directory for process execution in the client, while execution on the server writes the export to the workspace folder.
• Use explicit path exports only to the location defined in the Export Path field. There is no connection to the workspace and export is preserved after rollback or project deletion.

Export Item Name

Export item name in the export specification. Only available for Export mode > Static export item.

Export Item Variable

Variable containing the export item name in the export specification. Only available for Export mode > Dynamic export item.

Export Path

Modify to change the export file or directory. Default:

{:Workspc.OutputRoot}\results\{:Item.Name}\{:Project.Name}.v{:Project.Ver}.{:Ext}

Export Series

If set to yes, then multiple files per series will be exported, or additional columns will be created for table exports.

Confusion matrix

Input data

Select if an Image object level (defined in the domain) or a Flattened confusion matrix is used for the algorithm. Choose between:

• Image object level

• Flattened confusion matrix

Class filter

Select the ground truth classes. (Only available for Input data > Image object level.)

Ground truth

Select the source of the ground truth information. (Only available for Input data > Image object level.) Choose if the ground truth information is stored in a:

• Sample class

• Object variable

Input path

Select the path of an existing confusion matrix to be used as input for the computation. Only available for Input data > Flattened confusion matrix.

Add detailed statistics to export

If set to 'Yes' detailed statistics are included in the export, including User and Producer Accuracy, Hellden index and KIA per class.

Flatten confusion matrix

Export a single row of accuracy assessment values for each project instead of a standard confusion matrix. Only use if computing accuracy over multiple projects in a workspace.