Object Classification

Object Classification helps you identify what appears in images or video feeds. This guide explains how to set up and use a classification model in your project.

Input Options

You have multiple options for providing images to classify:

• Use your webcam (default setting)
• Connect a different camera by adding a custom camera to your setup
• Analyze pre-recorded pictures or videos by using the Media Loader element instead of Camera

Basic Setup

1. Open the Image Classification Inference element settings by clicking the ... button in the top right corner

   • Select a trained model using either the drop-down menu or the Select Directory button
   • Set the minimum confidence threshold for classifications

2. If you only want to view the classifications without saving, you can:

   • Delete the Image Inference Saver element
   • Click the Run button at the top of the screen
   • Open the Output Preview when the link appears

The first time you run this flow, dependencies will be installed automatically, which may take several minutes.

Saving Classification Results

If you want to save the classified images with annotations for later use:

1. Open the Image Inference Saver settings

2. Configure the following options:

   • Output Folder Path: Select where to save your results
   • Partition Name: Name the subfolder within your output directory
   • Image Format: Choose jpg, png, or npy

3. Select what to save:

   • Images
   • Annotations
   • Bounding boxes on the images
4. Set the minimum confidence threshold for saving results
5. Click Run and open the Output Preview when the link appears

The saved results can be used for further analysis or to improve future training datasets.