Train Your First Model

This section shows how to train and export an example deep learning model that can be used for defect segmentation. The data used for training is from an image dataset of Ethernet ports.


  1. The software running device should satisfy relevant system requirements. See the System Requirements section to check these requirements.

  2. Click here to download the image data and unzip the file.

When Mech-DLK is running, the resolution and scale of the screen should follow the correspondence in the table below. If the resolution and scale of your screen do not match those in the table below, some display problems may occur.

If you are using dual monitors, make sure that both have the same resolution and scale.



1280 × 800 (16:10)


1920 × 1080 (16:9)

100% or 125%

2560 × 1440 (16:9)

125% or 150%

3840 × 2160 (16:9)

150% or 175%

Training Process

flow chart
  1. Create a New Project: Click New Project in the interface, name the project, and select a directory to save the project.

    Do NOT include Chinese characters in the path name.
    new project
  2. Add the Defect Segmentation Module: Click icon create to add a module. Select Defect Segmentation and then click OK.

    add new module
  3. Import Image Data: Click Import/Export in the upper-left corner, select Import Folder and import the image folder you have downloaded.

    import images
    • Import Image: import one or more images.

    • Import Folder: Import all images from the selected folder.

    • Import Dataset: Import a dataset that contains images and corresponding labels. You can only import datasets in the DLKDB format (.dlkdb), which are datasets exported from Mech-DLK.

  4. Labeling: In this example, you will need to label the OK images and NG images in each dataset. OK means that the connectors meet quality requirements and NG means that there are defects such as deformations and fractures on the Ethernet ports. Labeling is to provide the information required by deep learning training.

    ok defect features

    For NG images, long press the left mouse button or right-click icon tool on the toolbar and then use the Brush Tool to select the regions with defects.

    label data1
    label data2

    Click icon eraser to use the eraser tool to remove the labeled region.

    For OK images, select an image and click the OK Label OK label tool on the labeling toolbar, or right-click the image and select Set to OK. The image will be labeled as OK. Please make sure that there is at least one OK image in each dataset.

    label ok
  5. Train the Model: On the Training tab,click Train in the lower-right corner of the interface to start training the model.

    training chart
  6. Validate the Model: After the training is completed, click  Validate to validate the model and check the results.

    result verification
  7. Export the Model: Click Export and select a directory to save the exported model (with file extension of .dlkpack).

    model files

Now, you have completed the training of the first model. Continue your reading to learn all the details of modules.

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.