Image Classification

Overview

Image Classification node predicts the most likely class label for the entire input frame using a loaded model.

Unlike object detection, this node classifies the frame as a whole rather than locating individual objects within it. Use this node when you need to categorize scenes (for example pass/fail, empty/occupied) or distinguish between broad image types.

Input

Input Image

image required

The image frame to classify. Connect this to a camera or upstream processing output.

Model Directory Path

string required

Path to the model directory to load for classification. See Model Directory Path for details.

Confidence Threshold

number required

Minimum confidence score for a label to be included in the output. Results below this value are suppressed. See Confidence Threshold for tuning guidance.

Ignore Labels

array

List of label strings to exclude from the output. Useful for suppressing catch-all or background labels that are rarely actionable.

Overlay Results

boolean required

When enabled, draws the top predicted label and confidence score as text on the output frame. See Overlay Results.

Advanced Setting

boolean

Reveals extra framework and architecture fields for non-.nam models. See Advanced Setting.

Output

Overlay Image

image

Output frame from the node. If overlays are enabled, the top predicted label and score are drawn as text on this frame.

Classification

object

Classification result object containing:

• label string: The predicted class label.
• confidence number: Confidence score for the prediction.