Green Classify

You can use the tool to identify and classify an object, or the entire scene in an image. The tool assigns a tag to the images, and uses the tag for sorting the images into classes. The tag is represented by a label, and each label has a percentage showing how confident the tool is in the assigned class.

You can use the tool to simply classify an object in an image, such as Part A, Part B, Part C, and so on. In addition, you can use the tool before other tools that perform inspections downstream in the tool chain. For example, the Green Classify tool can pass images of Part B to a Red Analyze tool downstream for further inspection. However, the Green Classify tool can pass images of Part C to a Blue Locate tool to count features instead.

You can also use the Green Classify tool downstream of a Red Analyze tool to classify the types of defects, or after a Blue Locate tool to classify the type of model that produced a particular view.

The Green Classify tool is available in two architectures:

• The focused architecture tool can quickly classify parts in most environments. The focused architecture uses feature sampling to learn the pixel information of images. You can set the magnitude of the feature sampler with the sampling parameters.

• The high detail architecture tool is more accurate than the focused architecture. The high detail architecture does not use a feature sampler because it samples the entire image. The training and processing time of the high detail architecture is longer than that of the focused architecture. The high detail architecture also uses a validation set. Moreover, you cannot assign multiple tags per view in high detail architecture. Each view has only a single corresponding tag and it means that non-exclusive mode is not supported.

Supported Features

Features	Green Classify Focused	Green Classify High Detail
View Inspector	Supported without heat map	Supported with heat map
Loss Inspector	Not supported	Supported
Validation Set	Not used in training	Used in training
Tool Parameters	Fewer parameters	More parameters for training and perturbation parameters, but no sampling parameters
Multi-class Classification (Non-Exclusive/Exclusive Mode)	Supported	Not supported
Resize Mode	Not supported	Supported