DEEP LEARNING
Deep Learning is the best-in-class deep learning-based image analysis software designed for factory automation.
Its field-tested algorithms are optimized specifically for machine vision, with a graphical user interface that simplifies neural network training without compromising performance.
Deep Learning solves complex applications that are too challenging for traditional machine vision, while providing a consistency and speed that aren’t possible with human inspection.
When combined with deep Learning rule-based vision libraries, automation engineers can easily choose the best the tool for the task at hand.
Deep learning toolset
Deep Learning tools are trained by example, unlike traditional rule-based vision algorithms. These tools are optimized for factory automation vision inspections and require smaller image sets for quicker training. The user-friendly GUI also provides a simple environment to manage and develop your applications. Choose between Blue Locate, Red Analyze, Green Classify, and Blue Read tools to solve applications that are too complex for traditional rule-based machine vision approaches.
Blue Locate for fixturing, counting, and assembly verification
The Blue Locate tool finds parts with variable appearance. It detects features on noisy backgrounds, in poorly lit environments, on low contrast parts, and even parts that flex or change shape. Blue Locate locates parts despite variations in perspective, orientation, luminance, glare and color by learning from the samples provided by the user.
Blue Locate is also a reliable solution for automated assembly verification. The tool can be trained to find a variety of components, even if they appear different or vary in size, to create an extensive component library. By creating layouts based on the product being inspected, the tool checks multiple feature locations and component types simultaneously, while adjusting to varying layouts.
Red Analyze for defect detection and segmentation
The Red Analyze tool finds subtle defects on a wide variety of part backgrounds and surface textures. By showing it examples of good and bad parts, it can be trained to tolerate normal variations in lighting and part positioning, while detecting flaws, contamination and other defects.
For situations where it’s not practical to collect defect images, or where the defects are highly inconsistent, Unsupervised mode can be trained from just good images and identify cases that deviate from the normal part appearance.
Red Analyze can also be used to segment specific variable areas in an image. These might be weld seams that are passed to a Green Classify tool; glue or paint regions whose coverage is then measured with traditional vision tools, or background features that are dynamically asked out of the image to simplify other inspections.
Green Classify for object and scene classification
The Green Classify tool is a robust classifier that can be used to distinguish between different types of objects, identify defect types, and even inspect images. Learning from a collection of labeled images, Green Classify identifies and sorts products into classes based on their common characteristics such as color, texture, materials, packaging, and defect type.
The tool tolerates natural deviation within the same class and reliably distinguishes acceptable variation from different classes. Green Classify solves complex classification tasks very quickly, eliminating complicated and time-consuming programming.
