OneBoxVision Blog

Setting up your inspection system to classify defects

When your inspection system is running, you need to be able to tell what types of defects the system is detecting.

OneBoxVision's machine vision library can be set up to classify defects based on measurements such as size,color intensity and aspect ratio.

Defect Types

The classifier in Imageflow is robust enough to accurately define and detect a variety of common defects in plastic inspection.

Our systems are typically set up to detect defects such as:

  • Carbons
  • Gels
  • Holes
  • Contamination
  • Tears
  • Wrinkles
  • Hazing

Accurate defect classification results in a more efficient inspection system, as users are not only able to detect the presence of defects, but also what specific types of defects were found.


Classification is Key

The key to any successful inspection system is classification. Comprehensive and accurate classification allows you to:

  • Identify re-occuring defects and take corrective action
  • Use reporting features more effectively
  • Make the inspection system more intuitive for operators.
  • Introduce alarms to the your inspection system when certain defects are found.
  • Utilize the IO by introducing rejection systems.

Tools for Engineers

Classification in Imageflow is done through Tool Manager and gives Engineers all the resources they need to find and classify defects in your material.

The classifier can be set-up in 4 simple steps.

  1. Record a video of the material to be inspected and load into Tool Manager
  2. Load the classifier tool.
  3. Find common defects in the material and view their measurements using blob analysis.
  4. Note these measurements and use them to apply the classifier.


An example of this process can be found in the video contained in this article.  


OneBoxVision are experts in machine vision and 100% inspections systems so we know the necessity of being able to accurately classify types of defects in your inspection.

This is why we have developed our machine vision applications to have intuitive and accurate defect detection that can lead to improved reporting and system performance.

Contact us to learn more >

Topics: Plastic film inspection Building vision networks