INDUSTRY

Semiconductor

In the semiconductor industry, quality control through a demanding production environment and detailed data analysis for each process determines the quality of the product.

The existing inspection method is once again inspected with the naked eye of the operator after the first inspection through the equipment to inspect irregular and unstructured defects.
However, with the artificial intelligence vision inspection solution, the front of the wafer can be fully inspected without the operator's review.

Experience the build-up of other companies that introduced LAON PEOPLE's artificial intelligence software + vision solution first.

INQUIRY
sales@laonpeople.com+82-1899-3058ONLINE INQUIRY
LIST

Product Specific Specifications

Information
SCROLL >>
Not registered post.
Wafer Crack Inspection Information

Applied Solutions


NAVI AI Mars

 

Wafers, the main material of semiconductors, are made of fragile materials, so they require a lot of attention in handling them.

​

Because microscopic cracks in wafers can easily break even with small shocks or temperature changes, vision inspection for cracks can improve quality and productivity.

 

 LAONPEOPLE uses AI to accurately detect crack that occurs atypical times​.






 

Overview of Inspection

  • ​Inspection Field: Semiconductor
  • Inspection Parts: Wafer
  • Inspection Type: Crack inspection
  • Configuration of Inspection Solution: NAVI AI Mars



Features and Benefits of LAON PEOPLE

  • ​Possible to conduct the vision inspection even without knowledge of image processing
  • Detect the location of atypical defects accurately

 

 

Crystal Diode Inspection Information
Applied Solution
LPMVC-EE400(GigE/4M) + NAVI SW
To examine the crack in a crystal diode with a traditional computer vision, advanced knowledge is required, and detection of defect may not be accurate. As shown in the image below, cracks and scratches are very difficult to handle with computer vision algorithms in terms of being unstructured at random locations. For example, if you try to detect a defect by obtaining an image that shows the difference between the original image and the inspection image, the residual image or noise processing in that image can be another problem.

However, applying the deep learning method makes it possible to perform high-performance inspection without difficulty. If you prepare an image of the defect from the user's point of view, the subsequent process can be left to deep learning, making it simple.

LAON PEOPLE has also tried inspection this case with computer vision technology and deep learning. Based on the review of 200 accepted/defective images, it took more than a month to implement them with computer vision technology and respond to exceptional situations with case by case, and it was not easy to satisfy coverage. However, using deep learning technology, we checked the results in less than a day, and the coverage was 100% as well.

Difficulties that have not been solved by conventional computer vision can be solved through deep learning. Deep learning detects foreign substances/cracks/scratches/stains that occur on the appearance or surface of the metal, can also detect stitches or cutting/sticking in the fabric, classify various parts, or read optical characters on a complex background. As such, deep learning can be an alternative to existing human visual inspection processes.


Overview of Inspection
  • Inspection Field: Semiconductor
  • Inspection Parts: Crystal diode (8 * 8mm)
  • Inspection Parts: Crystal diode (8 * 8mm)
  • Configuration of Inspection Solution
    - Inspection Camera: LAON PEOPLE LPMVC-EE500(GigE/5M)
    - Inspection S/W: LAON PEOPLE NAVI
    - Lens: 16mm
    - Lighting: Telecentric

 

Features and Benefits of LAON PEOPLE 
  • ​Possible to conduct the vision inspection even without knowledge of image processing
  • Accuracy at the level of inspection by human
  • Speed about 10 pages per second