Examples
Standard Software using Preferred Networks Visual Inspection
| Category | Electrical components |
|---|---|
| Workpiece | Screw |
| Inspection details | Visual Inspection |
| Application | Scratch inspection |
Before
This is an image of a defect on the surface of a screw.
In conventional image processing, it was necessary to set the detection area and the change in the brightness of the work surface in order to detect defects.
In conventional image processing, it was necessary to set the detection area and the change in the brightness of the work surface in order to detect defects.
After
By using deep learning, it is not necessary to make detailed settings, and it is possible to detect defective parts simply by registering an OK captured image.
[Main characteristics]
Inspection is possible by combining Rule-based image processing and Deep Learning.
With this software, you can easily realize processing such as image filter, alignment, image cropping processing by Rule-based image processing,
and then defect inspection by Deep Learning.
Furthermore, it is also possible to measure the area based on the rules for the inspection results of Deep Learning.
Rule-based image processing ・・・ VisionPro
Deep Learning image processing ・・・ Preferred Networks Visual Inspection
Suitable for the following inspections.
・ Defect detection
Inspection is possible by combining Rule-based image processing and Deep Learning.
With this software, you can easily realize processing such as image filter, alignment, image cropping processing by Rule-based image processing,
and then defect inspection by Deep Learning.
Furthermore, it is also possible to measure the area based on the rules for the inspection results of Deep Learning.
Rule-based image processing ・・・ VisionPro
Deep Learning image processing ・・・ Preferred Networks Visual Inspection
Suitable for the following inspections.
・ Defect detection