1. NS42 Smart Vision Sensor Product Reference Guide
  2. Using Machine Vision Tools
  3. Using Anomaly Detection
  • NS42 Smart Vision Scanner Product Reference Guide
Using Anomaly Detection

Using Anomaly Detection

Aurora Deep Learning Editor is required to run Anomaly Detection in Aurora Focus.
Anomaly Detection is available on FS24 and NS42 devices only.
  1. Add the
    Anomaly Detection
     tool from
    Presence/Absence
     tools list.
  2. Select a model from the drop-down menu and click
    Edit
     to edit the model in Aurora Deep Learning Editor or click
    Load
     to implement a specific model type.
    anomaly detection choose model
  3. View Results:
    1. The score is a numeric representation of how flawed the processed image is when compared to the reference model. A higher score indicates a greater number of defects.
    2. Threshold values can be set to define what scores are considered as pass or fail with high or low confidence.
    3. Heatmap provides a graphical representation of the anomalies. The higher the color intensity (according to the chosen palette), the stronger the anomaly detected in the image.
    4. The Scale parameter is used to adjust the intensity of colors when an anomaly is detected.
    anomaly detection results aurora focus