Identify, count, and classify parts based on their appearance
Slight variations in appearance can cause complications for an automated inspection system. For certain identification, counting, and classification applications, human inspectors and traditional machine vision solutions are not suitable. This is the case for many kitting applications, where the inspection system must be able to identify, count, and classify various different items pre-assembly.
Cognex deep learning-based image analysis excels at kitting inspections because it is able to generalize the distinguishing features of an item such as size, shape, and surface features based on training images. The ViDi Blue-Locate Tool identifies and counts parts in a package. The ViDi Green-Classify tool uses the deep learning-based model to classify the items by category. Taken together, this information confirms that the right items will be included in a multi-part unit.