CG method (cross grid method) Automatic Visual inspection software CrossVector

Attribute:

A completely new automatic visual inspection package software with unique algorithms

Share

Share

A completely new automated visual inspection package software with unique algorithms

 

A completely new visual inspection software that looks like AI, but isn't AI

After registering at least 10 images of good products, you can start inspection right away!
If an overdetection occurs, you can register it right away!
Differences from good products are detected as deformations in the shape and differences in vectors!

The camera is compatible with the CS series.
We recommend using a global shutter of USB camera.

 

New visual inspection algorithm

Going beyond just brightness, it is now possible to detect the difference between a good product and a defective product by looking at the shape and vector differences!

Good product imageNG detection image

 

When inspecting the above objects, detection was unstable due to the balance of light when using conventional rule-based types. However, by using the CG method (cross grid method), stable detection is possible by simply memorizing several dozen images of non-defective products.

There is no need for huge amounts of training data or long learning periods as with conventional AI.

 

Strengths of the CG method (cross grid method)

Like AI
you can start by registering 10 or more quality images.
It's not AI
so you don't need a high spec PC and it only takes a few minutes to learn.
New inspection technology
Excellent for detecting small scratches and dents that occur on the edges of workpieces

 

Inspection algorithm concept
The CG method (cross grid method) registers a good product, memorizes the shape of the image of the good product, and compares the position and direction of each edge that makes up the contour.
This makes it possible to detect vectors that are not present in the good product.

 

 

Frequently asked questions about automatic visual inspection
Example of an answer when using the CG method (cross grid method)
 
Q:What is the detection accuracy?
A:Events can be detected from as little as 3 pixels, so even subtle changes may be detectable.
 
Q:What about the stability and reproducibility of the test?
A:Even if there are lighting conditions or individual differences, it remains relatively stable because it looks at vectors.
 
Q:Is it easy to learn and set up?
A:It's very easy; just take several dozen photos of good quality products and press the learning button.
 
Q:How long does one inspection take?
A:It depends on the resolution of the camera used, but it is generally within 0.5 seconds.
 
Q:What kind of tests is it effective for?
A:It is suitable for detecting defects that occur in finished products with a fixed shape.
 Since the test involves memorizing shapes, it has difficulty with objects that are irregular in shape or have different positions and orientations.
 
Q:Can it be linked to other systems?
A:Since I/O control is possible, it is possible to start inspection and output OK/NG inspection results.

 

Automatic visual inspection: 3 easy steps to implementation
No need to worry if you don't have image processing knowledge in-house!
Step 1: Interview and sample confirmation of test subjects
 
We ask you "What kind of products do you want to inspect and how?" and propose the optimal inspection method based on the samples.

 

For example, we check for scratches, whether there is a label, whether there is a chip, etc.

We accept photos and samples of the object to be inspected and the shape of the workpiece.
Click here for more information.
 

Step 2: Propose the optimal system configuration and lend a demo unit
 
We will propose a system that combines lighting, cameras, and appearance inspection software based on the samples we receive, and you can use the actual demo machine to check the detection accuracy, etc.
 
Select lighting and lenses to suit your workpiece
Explain how to use visual inspection software
 

Step 3: Introduction
The operation is simple and it can be used immediately in the field.

  Support for registering test items and setting judgment criteria via telephone, web conference, etc.

 

If you have any questions, please feel free to consult our technical staff at any time.

This website uses cookies for best user experience, to find out more you can go to our Privacy Policy  and  Cookies Policy