Samples Category
Instructions for Use
The Samples Category group in the Training Data Cleaning tab provides a set of efficient shortcuts and buttons, allowing users to perform rapid, batch status categorization for the currently browsed sample. This operation automatically checks if the label conforms to the classification category. For example, if marked as All Positive Samples, the system will determine whether the label contains pixels other than All Positive Samples. If it does, the label will be modified to All Positive Samples; if not, no changes will be made. This is suitable for macro-level quality assessment and filtering of entire samples.
Available starting from SuperMap iDesktopX 2026 version.
Operating Instructions
- Select the target sample: In the Training Data Manager, expand the Unclassified group and select a sample record. The map will synchronously display the imagery and label for that sample.
- Perform the categorization operation: Click the corresponding button under the Samples Category group, or use its shortcut key, to mark the current sample as the corresponding category. The provided function descriptions are as follows:
- All Positive Samples [A]: Marks the current sample as a completely correct positive sample. Suitable for high-quality positive samples where all targets in the entire image are correctly labeled with no omissions.
- All Negative Samples [S]: Marks the current sample as a negative sample (background or no target). Suitable for samples that contain no targets throughout the entire image, or serve as pure background.
- Delete [D]: Moves the current sample into the Deleted group. Suitable for samples with extremely poor labeling quality, unusable samples, or erroneous samples.
- Do Not Modify [F]: Maintains the sample's current state without any changes. Suitable for samples that have been reviewed and deemed to require no modification, explicitly marking them to avoid duplicate checks.
- Undo [Ctrl+Z] / Redo [Ctrl+Y]: Undoes or redoes the previous categorization operation.
- View the result: After the operation is completed, the sample record will be moved from the Unclassified group to the corresponding All Positive Samples group, and this operation will be recorded in the *.sda.xml file, not in the Modified folder of the storage path. Simultaneously, the next record in the Unclassified group will be selected in the Training Data Manager.
Label Editing
Instructions for Use
The Label Editing group in the Training Data Cleaning tab provides a set of vector drawing tools, allowing users to visually modify sample labels, including adding positive sample regions, subtracting negative sample regions, and adjusting region shapes. This operation generates new label files and is a core method for microscopically correcting samples.
You might use this function when you find local errors in a sample's label (such as over-labeling, under-labeling, or inaccurate boundaries).
Available starting from SuperMap iDesktopX 2026 version.
Operating Instructions
- Select a sample whose label needs to be modified in the Training Data Manager.
- Under the Label Editing group, select the desired tool and complete the editing operation:
- Positive Label: Draw a polygon on the image in the map to add a positive sample region.
- Negative Label: Draw a polygon intersecting with existing positive labels on the image in the map to subtract that region (marking it as a negative sample/background).
- Edit Node: First, click to select an existing polygon on the map, then use this function to add, delete, or drag nodes to adjust its shape.
- After editing is complete, click the Save button. The system will update and generate a new label file based on the drawn positive and negative label regions. The new label file will be auto-saved to the Modified folder under the project's storage path. The sample's group in the Training Data Manager will automatically be categorized as Modified.
Related Topics