Image Classification

The Image Classification component enables the user to classify images into various classes also providing accuracy for the same.

Setup

  1. Make sure Intellibot Machine Learning Plugin is installed. Refer Installing Plugins documentation for details.

  2. In Toolbox, expand Machine Learning. Drag and drop Image Classification connector to Global Objects.

  3. Save the project on your machine. The ML model will be generated in the same directory.

Configuration

  1. To configure Image Classifier, expand Global Objects, right click on ImageClassifier and select Configure.

  2. The Image Classifier Model screen opens. User can add Image Classes as per the project requirements by clicking on the plus sign next to the Image Class. For example, let’s add healthy and diseased as class names. Both the image classes will be available to view under dropdown menu of Image Class.

    Screenshot

  3. After the Image Class is created, select each image class one-by one and click on Import button to add images for that class. Alternatively, user can capture images from the screen using the Capture button.

    Screenshot Screenshot

  4. Once the images for all classes are uploaded, the next step is to train the model by clicking on Train section of the Image Classification Model Designer.

Training the Model

  1. The Train tab will expose the Classifier Settings which the user can change as per the requirements. Alternatively, the user can train the model with the default settings.

  2. Click on Train. After the model is trained, it will show the model accuracy. The user will be able to see the image classification of test data with respective confidence levels.

    Screenshot

  3. Depending on the model accuracy, the user can decide to tweak the Classifier Settings and retrain the model.

Testing the Model

  1. Click on Test tab to test the created model with test data.

  2. Click on Import/Capture button to add the test image(s).

  3. Click on Test button. The classification result will be displayed under the image.

    Screenshot

  4. Close the Image Classification Model Designer.

Using the Model

  1. Double click on the ImageClassifier under Global Objects to expose the various components under Methods in Object Explorer.

  2. Drag and drop the ClassifyImage component to the Design Surface which takes an image as an input and returns the class and confidence predicted by the model.

  3. Connect this component to your automation flow.