Detects anomalies and aesthetic defects
The ViDi Red-Analyze tool is used to quickly identify defects on complex parts and surfaces. Using a sample set of good images and bad images with labeled defects, ViDi Red-Analyze learns the normal part variation, while creating a comprehensive understanding of the defects.
Identifies unpredictable defects
For situations where it is difficult to collect images of defects, or if the failure modes are yet unknown, ViDi Red-Analyze can learn from just good images in Unsupervised mode. With Unsupervised mode, the tool learns the normal condition and identifies areas of the images that stray from this normal appearance.
Segments images to identify areas of interest
ViDi Red-Analyze can also be used to segment specific areas of an image. By teaching it areas of interest across the sample set, the tool learns to identify and highlight these areas. This can be used as a dynamic region of interest or mask, providing a unique method to simplify the deep learning application.
Eliminates complicated, time-consuming programming
The ViDi Red-Analyze tool drastically reduces development and inspection time and solves random defect detection applications that are challenging to program with traditional machine vision. The software’s intuitive, easy-to-use interface makes it simple to set up and deploy jobs, directly on the factory floor.