Accurately classify and sort electronic capacitors within a single image using deep learning
Graphical programming environment for deep learning-based industrial image analysis
Classifying electronic components can be especially challenging when parts fall into multiple classes, each with some visual variation. Capacitors vary in type (ceramic and electric) and also by size and color, depending on their manufacturer and specifications. Even within the same type, there can be confusing variations in pattern. Their cylindrical shape and lighting can add even more complexity. VisionPro Deep Learning offers a deep learning-based alternative to automate multiple classifications within a single image.
Using the defect detection tool, an engineer trains the software in supervised mode on a set of annotated images where both gold and electric capacitors are categorized as “good” parts. During runtime, the model extracts and segments all electric and gold capacitors as one type. In the second part of the inspection, the classification tool learns the attributes of each capacitor, while tolerating variation within the same type. In this way, it can distinguish different electric capacitors by their color and marking, even though they look visually similar. Based on the model developed during training, Cognex Deep Learning accurately classifies and sorts capacitors within a single image during runtime.