• Image Processing
• Basics of Image Processing
• Principle of Image Processing
• Inspection Functions
• Feature Description

• Explanation of Functions
• Peripheral Equipment
• Area Judgment by Binarization
• Position Detection
• Dimension Measurement

• Ask the Experts
• Price Info
• Free Trial Unit
• Downloads
• Contact Us

Home > Technology > Inspection Functions

Ask the ExpertPrice InfoTrial UnitCatalogPrint Page

Machine Vision Technology

---

Inspection Functions

Area sensor (binary conversion)

This function detects the area by counting the number of white or black pixels within the measurement window. With color processing, the area is detected by counting the number of extracted or unextracted color pixels.

Dimension measurement

Various dimensions are measured by setting a window on the image and measuring the distance between the edges detected using the window.

Inner diameter measurement

Outer diameter measurement

Blob

Regarding the amount of a characteristic

• Barycentric position: A barycentric position is measured using X and Y coordinates.	• Area The number of pixels in a blob area is calculated.
• Angle of the main coordinate: The angle of the main coordinate to the horizontal coordinate (X-coordinate) is measured up to ±90º.	• Feret: The lengths of the X and Y coordinates of the circumscribed rectangular of a blob is measured.
• Perimeter length: The perimiter of the blob is measured.	• Degree of circularity The degree of circularity of a blob is calculated based on the degree of perfect circularity, 1.000.

A collection of pixels in a binary image (black and white) that have identical density (0 or 255) is called a "blob". The blob mode measures the amount of characteristics of a blob. The characteristics that can be measured include the area, the barycentric position and others.

Stain Detection

Detection principles (with detection in one direction)

1. Measures an average concentration while shifting a small area (segment) of any size in the area by the distance of 1/4 of the segment size.
2. Measures the difference between the maximum concentration and minimum concentration in the 4-segment range of the detection direction, including the standard segment ( A: 95 in the right figure). This value serves as the "stain level" of the standard segment.
3. If the stain level exceeds the set threshold, the standard segment is counted as a stain. This count value serves as the inspection result called "stain amount."
   After that, A to C are repeated while shifting the standard segment in the area by a set distance.
   
   
   
   
   

Benefits of the stain detection algorithm when compared to general techniques that use binary settings

• The stain detection algorithm does not need custom binary settings tailored to a workpiece because it processes the image contrast.
• The stain detection algorithm is strong against illumination change because it judges stains based on the change in contrast, not the absolute value.

Trend edge position

The trend edge function scans in a specified direction within a measurement area and calculates the minimum, maximum and average position of each point. This function is effective for measuring maximum and minimum outer diameters and for detecting flaws.

• Highly accurate detection of position: Reduce the segment size.
• Reduction of processing time: Reduce the shifting width of a segment (= moving distance)
• The trend direction is the direction in which the segment moves.

Intensity check

KEYENCE sensors enable measurements of average maximum, minimum and deviation values of intensity (brightness) in a measurement area. This feature allows KEYENCE sensors to be used not only for presence or absence checks but also for measurement of large variations of illumination and for high speed flaw detection.

Position Detection

Pattern Recognition by Gray Processing

In the case of gray processing, an image is treated as a gray scale picture of 256 tones. Unlike binarization which only recognizes black and white, gray processing uses all the information of each pixel which is divided into 256 tones and thus, the target object's pattern shape is accurately recognized.

Recognition of Coordinates on the CCD Image

Because an image projected on the CCD pickup element is divided into pixels of 480(height) × 512(length) as shown in the figure, the X-Y coordinate of any shape within the image is obtained within the range of 0, 0 to 511,479.

Inspection examples

Robot Handling (Detection of Absolute Position)

Move the robot so that the hole of the target object matches the home position of the screen and then pin the hole.