Engineering KnowledgeEnglish日本語 |
->

Photoelectric Sensors Laser
Sensors		 Pressure Sensors Safety
Curtain		 Displacement Sensors Optical Micrometers Digital Microscopes Laser Markers PLC
Units		 Barcode Readers
 
Ask the ExpertExpert Price InfoPrice Trial UnitUnit Catalog DownloadCatalog
Your guide to Machine Vision
Guide Topics
Machine Vision
System parts
Machine Vision Systems
Vision Techniques
Lighting Techniques
Major Illumination Types
Illumination Intensity
Illumination Basics
Specific Lighting Techniques
Polarizing Filters
Information About Lenses
Lens Basics
Lens Selection
Lens Characteristics
Image Processing Techniques
Major Image Processing Methods
Grayscale Edge Detection
Normalized Correlation Search
Binarization Stabilization Technique
Binarization Application Examples
Labeling
COUNT
Examples of Technique Application
Coated Glossy Surface, Engraved Mark Inspection
Detection of Scratches and Chips on a Low Contrast Surface
Detection of the Dimensions and Shape of Transparent Objects
Detection of Concave and Convex, Print on Curved Surface, and Contamination
KEYENCE General Catalog

Keyence Product Information Site
KEYENCE Machine Vision product information

Technical Glossary

World Standards Information
Machine Vision Systems
  

Grayscale Edge Detection
Our vision system comes with a grayscale edge detection mode, enabling position detection and dimensional inspection. Unlike binary edge detection, grayscale edge detection works with changing ambient illumination intensity and provides sub-pixel processing.
Principle of Edge Detection
By setting the window in an X direction and averaging the pixel data in a Y (vertical) direction for every row, changes in concentration can be plotted. Assuming that the brightest and darkest data of this graph represent 100 and 0 respectively, an edge recognition threshold can be set within this range. If the threshold set is too low, small changes in brightness will also be recognized as an edge, while setting it too high, close to 100, means that small changes in brightness are not recognized as an edge.
Set-up for edge detection   Setting a detection window for edge detection   Edge detection output

Edges data with a high threshold setting Edges detected with a high threshold setting
 
Edge data with a lower threshold setting Edges detected with a low threshold setting
Unaffected by Changes in Illumination Intensity
Even if the brightness of the surrounding environment darkens, grayscale edge detection recognizes the variation of brightness from a scale of 0-100 and adjusts with these changes.
Bright image   Dark image
 
Data from bright image   Data from dark image
Sub-pixel Processing: Up to 0.1 pixels can be Detected
Every piece of gradation for every pixel is analyzed, allowing for sub-pixel processing from 0.1-0.001 pixels.
Variation of shades   Regular processing   Sub-pixel processing


More on this topic...
-> A Desktop Reference for Sensors, Vision Systems and Measuring Instruments. General Catalog
-> The latest installment from the 50 Key CV solutions details everything about edge detection. The latest installment from the 50 Key CV solutions details everything about edge detection.
Have confidence in your vision systems performance Have confidence in your vision systems performance
Click here to ask our Machine Vision experts a support question.
Need price information? Click Here
For information on a trial unit click here.

Page Top
Copyright(c) 2010 KEYENCE CORPORATION. All rights reserved. | Privacy | Terms of Use | Contact | Produced by Produced by Keyence