Use Labeling Tools

Mech-DLK provides a variety of image labeling tools for different algorithm modules. This topic introduces how to use these labeling tools to label targets in images, providing the necessary information for training deep learning models.

Label Tool

OK Label

Supported modules: Defect Segmentation, Unsupervised Segmentation

Use the OK Label to label an image as OK.

NG Label

Supported module: Unsupervised Segmentation

Use the NG Label to label an image as NG.

Labeling Tool

Polygon Tool

Supported modules: Instance Segmentation, Defect Segmentation, Object Detection, Text Detection, Fast Positioning

The Polygon Tool can draw polygon labels with more vertices, which is suitable for targets of complex shapes or those near the image edges.

  1. Click tools introduction polygon icon (or press P on the keyboard).

  2. Click the first position (vertex) in the selection region, then click the second one, third one, etc., to draw the labels, and right-click to finish. (At least three vertices are required.)

  3. If multiple label classes are created, colors corresponding to different label classes should be selected.

After labeling, use the Select Tool to select a label and adjust the label by the following methods.

  • Click the label edges to increase the number of vertices.

  • Right-click a vertex to delete it.

  • Long press the left mouse button and drag the vertex in any direction to modify the label shape.

Ellipse Tool

Supported modules: Instance Segmentation, Object Detection, Fast Positioning

Use more vertices to make elliptical selections. This tool is suitable for elliptical objects.

  1. Right-click tools introduction polygon icon and then click tools introduction oval icon (or press L on the keyboard).

  2. Click the first position (vertex) in the selection region, and then continue clicking. An elliptical label should have at least six vertices.

  3. If multiple label classes are created, colors corresponding to different label classes should be selected.

After labeling, use the Select Tool to select the label and then long press the left mouse button to drag the vertex in any direction and thus modify the label shape.

Rectangle Tool

Supported modules: Instance Segmentation, Object Detection, Fast Positioning

The Rectangle Tool can be used to draw rectangular labels, which is suitable for rectangular objects.

  1. Right-click tools introduction polygon icon and then click tools introduction rectangle icon (or press R on the keyboard).

  2. Long press the left mouse button in the selection region, move it in any direction, and then release the left mouse button to finish the rectangular selection.

  3. If multiple label classes are created, colors corresponding to different label classes should be selected.

Free Rectangle Tool

Supported module: Text Detection

The Free Rectangle Tool can be used to draw a rectangle around the text area. Therefore, it is recommended to select this tool for rectangular text areas.

free rectangle tool

  1. Click tools introduction free rectangle icon (or press K on the keyboard).

  2. Move the crosshair cursor in the selection region and left-click to select the starting point of the area you want to outline. Then, move the cursor and left-click again to select the second point to define the range you want to outline. At this point, a straight line will be created, forming one side of the rectangular selection frame.

  3. Moving the cursor to either side of the straight line will allow you to create a rectangle. Click the left mouse button to complete the labeling.

During the process, you can click the right mouse button to cancel the labeling.

Brush Tool

Supported module: Defect Segmentation

Use the Brush Tool to draw labels of any shape. This tool is suitable for defects with complex shapes.

tools introduction pen

  1. Right-click tools introduction polygon icon and then click tools introduction pen icon (or press B on the keyboard).

  2. Adjust the slider to set the thickness of lines according to the size of defects.

  3. Long press the left mouse button in the selection area, move in any direction, and then release the left mouse button to finish the drawing.

Autofill Lasso Tool

Supported module: Defect Segmentation

Draw arbitrarily shaped labels by forming closed shapes with brush paths. This tool is used for defects with complex shapes.

tools introduction lasso

  1. Right-click tools introduction polygon icon and then click tools introduction lasso icon (or press N on the keyboard).

  2. Long press the left mouse button in the selection area and move in any direction.

Smart Labeling Tool

Supported modules: Instance Segmentation, Defect Segmentation, Object Detection, Fast Positioning

The Smart Labeling Tool can be used to automatically select the objects in the image.

When multiple objects in an image have large color differences and are scattered, you can use the Smart Labeling Tool to conveniently label the objects in the image.

  1. Click tools introduction smart labeling icon (or press A on the keyboard).

  2. Move the cursor in the selection region and then click the object to be labeled.

    • If the selection cannot completely cover the object, click the uncovered part to expand the selection area.

    • If the selection covers the areas outside the object, right-click these areas to reduce the selection area.

  3. Click Apply in the upper-left corner of the selection region.

You can use the Select Tool to fine-tune the labeled contour by the following steps:

  1. Use the Select Tool to select the label to be adjusted.

  2. Adjust the contour in one of the following three ways according to actual situation. Please ensure that the selected area closely aligns with the object contour.

    1. Place the mouse cursor on a vertex of the contour. When the cursor turns into tools introduction smart labeling change2, long-press the left mouse button and drag the vertex to adjust the contour.

    2. Place the mouse cursor on a vertex of the contour. When the cursor turns into tools introduction smart labeling change2, click the right mouse button to delete the vertex.

    3. Place the mouse cursor on the contour. When the cursor turns into tools introduction smart labeling change1, click the left mouse button to add a vertex.

When the objects differ greatly in colors and have clear contours, it is recommended that you label multiple objects at a time and click Apply. If the objects are not obviously distinct, it is recommended to label one at a time.

Currently, the Smart Labeling Tool can work in two modes, and it will automatically switch between them according to the actual situation. Users do not need to manually select a mode.

  • Standard mode (GPU): In this mode, the Smart Labeling Tool has robust labeling capabilities. This tool will automatically switch to the standard mode if the running device has a GPU, with available GPU memory of at least 2.5 GB, and there is no project undergoing training or validation.

  • Lite mode (CPU): In this mode, the Smart Labeling Tool has basic labeling capabilities. This tool will automatically switch to the lite mode if the running device has no GPU, or the available GPU memory is less than 2.5 GB, or there is a project undergoing training or validation.

Pre-trained Labeling Tool

Supported modules: Instance Segmentation, Defect Segmentation, Classification, Object Detection, Text Detection

After you validate a model, you can import new image data to the current module and use the pre-trained labeling feature to perform auto-labeling based on this model.

The pre-trained labeling feature is available only when the current module contains validated models.

The Pre-trained Labeling Tool can be used only on the following three types of data:

  • Unlabeled data

  • Automatically labeled data (images with a yellow triangle at the front of the image number)

  • Manually fine-tuned data after automatic labeling (images with a yellow triangle at the front of the image number)

    The Pre-trained Labeling Tool can only label images that are not manually labeled. If you need to use the Pre-trained Labeling Tool on manually labeled images, you must clear these labels first.
    The pre-trained labeling feature can outstandingly reduce the cost of manual labeling. However, the accuracy of the results depends on the model that you use. We recommend that you train a high precision model before you use the pre-trained labeling feature.

You can use the Pre-trained Labeling Tool in one of the following three methods.

  • Method 1: Pre-trained labeling button

    1. On the upper part of the image list, click the Pre-trained labeling button. All images in the image list will be labeled by using the Pre-trained Labeling Tool.

    2. After the labeling process is finished, you can view a yellow triangle in the upper left corner of the sequence number of the labeled images.

  • Method 2: Pre-trained labeling option

    1. In the image list, select one or more images that you need to label.

    2. Right-click the images and select Pre-trained labeling, the labeling process starts. After the labeling process is finished, you can view a yellow triangle in the upper left corner of the sequence number of the labeled images.

  • Method 3: Pre-trained Labeling Tool

    1. In the labeling toolbar, right-click tools introduction smart labeling icon, then click tools pre labeling tool icon (or press D on the keyboard) to select the Pre-trained Labeling Tool.

    2. On the upper part of the image, click Start labeling, and the Pre-trained Labeling Tool will start to label the image. After the labeling process is finished, you can view a yellow triangle in the upper left corner of the sequence number of the image.

A yellow triangle will be displayed at the front of the sequence number of each labeled image. If you fine-tune the results of pre-trained labeling, these images are still considered as pre-trained labeled, and the yellow triangles do not disappear.

  • If you clear all labels of an image, the yellow triangle before it will be cleared.

  • In the Classification module, if you modify the class labels of an image, the yellow triangle before it will be cleared.

VFM Labeling Tool

Supported modules: Instance Segmentation, Object Detection, Text Detection

Mech-DLK incorporates the visual foundation models (VFMs) developed by Mech-Mind. You can use the VFM Labeling Tool powered by the visual foundation models to quickly label datasets.

The VFM Labeling Tool can be used only on the following three types of data:

  • Unlabeled data

  • Automatically labeled data (images with a yellow triangle at the front of the image number)

  • Manually fine-tuned data after automatic labeling (images with a yellow triangle at the front of the image number)

  • To use the VFM Labeling Tool, the available GPU memory on the current device must be greater than 6 GB; otherwise, the model conversion may fail.

  • The VFM Labeling Tool can only label images that are not manually labeled. If you need to use the VFM Labeling Tool on manually labeled images, you must clear these labels first.

You can use the VFM Labeling Tool in one of the following three methods.

  • Method 1: VFM labeling button

    1. On the upper part of the image list, click the VFM labeling button. All images in the image list will be VFM-labeled.

    2. After the labeling process is finished, you can view a yellow triangle in the upper left corner of the sequence number of the labeled images.

  • Method 2: VFM labeling option

    1. In the image list, select one or more images that you need to label.

    2. Right-click the images and select VFM labeling, the labeling process starts. After the labeling process is finished, you can view a yellow triangle in the upper left corner of the sequence number of the labeled images.

  • Method 3: VFM Labeling Tool

    1. In the labeling toolbar, right-click tools introduction smart labeling icon, then click tools super labeling tool icon (or press S on the keyboard) to select the VFM Labeling Tool.

    2. On the upper part of the image, click Start labeling, and the VFM Labeling Tool will start to label the image. After the labeling process is finished, you can view a yellow triangle in the upper left corner of the sequence number of the image.

After the labeling is completed, Mech-DLK will generate a label class named "model_label" for the labels created by the visual foundation models.

Mask Tool

If there are some irrelevant parts that may interfere with model training/inference, you can use the Mask Tool to cover such parts. The masked parts will not be involved in training/inference.

You can choose among the following three mask tools built in the software according to actual needs.

  • The Mask Type includes Mask single image and Mask globally.

    • Mask single image: the mask is only valid in the current image.

    • Mask globally: the mask is valid in all images.

  • It is recommended to set a distinct color for the mask.

Mask Polygon Tool

Supported modules: Instance Segmentation, Unsupervised Segmentation, Defect Segmentation, Object Detection, Fast Positioning

  1. Click tools introduction mask polygon icon (or press Shift + P on the keyboard)

  2. Set Mask Type and Mask fill.

  3. Click the first position (vertex) in the selection area, then click the second one, third one, etc., to draw the labels, and right-click to finish.

Mask Brush Tool

Supported modules: Instance Segmentation, Unsupervised Segmentation, Defect Segmentation, Object Detection, Fast Positioning

  1. Right-click tools introduction mask polygon icon and then click tools introduction mask pen icon (or press Shift + B on the keyboard)

  2. Set Mask Type, Mask fill, and Brush size.

  3. Long press the left mouse button in the selection area, move in any direction, and then release the left mouse button to finish the drawing.

Mask Lasso Tool

Supported modules: Instance Segmentation, Unsupervised Segmentation, Defect Segmentation, Object Detection, Fast Positioning

  1. Right-click tools introduction mask polygon icon and then click tools introduction mask lasso icon (or press Shift + N on the keyboard)

  2. Set Mask Type and Mask fill.

  3. Long press the left mouse button in the selection area and move in any direction.

Eraser Tool

Labeling Eraser Tool

Supported module: Defect Segmentation

The Labeling Eraser Tool can be used to erase the labeled region.

  1. Click tools introduction eraser icon (or press E on the keyboard)

  2. Long press the left mouse button in the selection area and move in any direction.

    Adjust the slider to change the eraser size.

Mask Eraser Tool

Supported modules: Instance Segmentation, Unsupervised Segmentation, Defect Segmentation, Object Detection, Fast Positioning

The Mask Eraser Tool can be used to erase the masks.

  1. Right-click tools introduction eraser icon and then click tools introduction mask eraser icon (or press Shift + E on the keyboard)

  2. Long press the left mouse button in the selection area and move in any direction.

    Adjust the slider to change the eraser size.

Grid Tool

Grid Cutting Tool

Supported module: Defect Segmentation

In industrial inspection scenarios, if the size of images captured by the camera is large, smaller defects may be inconspicuous. If training in such a case is performed, defects are difficult to detect. You can use the Grid Cutting Tool to cut the large images into cell images of the same size according to the set dimension. Defect labeling should be completed for all images before the application of this tool.

  1. Click tools introduction grid cut icon (or press U on the keyboard).

  2. Set Rows and Columns and then click Apply.

    • Place the cursor in the parameter boxes and then scroll the mouse wheel.

    • Enter values in the parameter boxes.

      Note that the number of rows and columns should not be too large, or else the number of cell images after cutting is great, which slows down subsequent inference.

Grid Selection Tool

Supported module: Defect Segmentation

Right-click tools introduction grid cut icon and then click tools introduction grid select icon (or press I on the keyboard) to open the Grid Selection Tool. By default, the cell images with defect labels are checked, which will be added into the training/validation set. You can select cell images with and without defects on demand. Click the Preview button in the upper right corner of the selected image to preview the cell images.

  • Select defects: Select all the cell images containing defects to put into the training/validation set.

  • Select all: Select all the produced cell images into the training/validation set, and set those containing no defects to OK.

  • Clear selection: Clear the selections on cell images.

Text Recognition Tool

Supported module: Text Recognition

The Text Recognition Tool can be used to define the text recognition range and automatically generate the recognition result.

text recognition tool

  1. Click tools introduction text recognition icon (or press J on the keyboard).

  2. Long press the left mouse button in the selection region and move the cursor to draw a rectangle. Release the mouse button to confirm the selection.

  3. The automatically recognized characters will be displayed right under the selection frame. If the recognition result is correct, click OK to confirm it; otherwise, please correct the result first and then click the OK button.

If Unconfirmed is displayed right under the selection frame, click the frame and then confirm the recognition result.

Template Tool

Supported modules: Instance Segmentation, Object Detection, Text Detection

You can use the Template Tool to set an existing selection as a template. After setting, you can use this template to rapidly label contours and objects with the same pose.

It is suitable for scenarios where there are multiple neatly arranged objects of the same type in an image, which can improve labeling efficiency.

tools introduction seal

  1. Click tools introduction seal icon (or press C on the keyboard).

  2. Click the region that needs to be set as the template.

  3. Move the template to the object that you need to select, adjust the angle of the template to make it fit the object, and then click it.

    • Coarse adjustment: Press and hold the Shift key and then scroll the mouse wheel.

    • Fine adjustment: Adjust the Rotation angle parameter.

During labeling, press and hold the Ctrl key and click the selection to switch the template. You can also click Replace template and then click the selection to achieve the same purpose.

ROI Tool

Supported modules: Instance Segmentation, Defect Segmentation, Classification, Object Detection, Fast Positioning, Text Detection, Text Recognition, Unsupervised Segmentation

You can use the ROI Tool to set the region of interest.

Setting the ROI can avoid interferences from the background.

  1. Click tools introduction roi icon (or press O on the keyboard).

  2. Adjust the ROI frame in the selection region.

  3. Click tools introduction OK to the lower-right corner of the ROI to apply the current ROI or click tools introduction cancel to disable the ROI Tool.

  4. Click the Reset button in the upper left corner of the image to reset the ROI.

Select Tool

Supported modules: Instance Segmentation, Defect Segmentation, Classification, Object Detection, Fast Positioning, Text Detection, Text Recognition, Unsupervised Segmentation

You can use the Select Tool to select, move, and adjust the selections.

  1. In the labeling toolbar, click tools introduction select icon (or press F on the keyboard) to select the Select Tool.

  2. Place the cursor in the labeling area of the image and use the mouse wheel to scroll forward to zoom in and scroll backward to zoom out.

  3. Move the cursor in the selection region and then click the selection to be processed. Select multiple selections by pressing and holding the Ctrl key.

  4. Right-click and drag in the labeling area to select all labels within the selected area.

Adjust size

tools introduction select1

Rotation

tools introduction select2

Move

tools introduction select3

We Value Your Privacy

We use cookies to provide you with the best possible experience on our website. By continuing to use the site, you acknowledge that you agree to the use of cookies. If you decline, a single cookie will be used to ensure you're not tracked or remembered when you visit this website.

  • Preference
  • I accept
  • I don't accept