You are here: Start » Program Examples » Washers

Washers

Aim

This example shows how to find washers with defects or not properly centered.

Input

Set of images with multiple washers.

Output

Image with marked broken and not properly centered washers.

Histogram of distances between inner and outer contour of a washer.

Hints

Washers can be located by thresholding image and splitting resulting region into blobs.

The FitCircleToStripe filter can be used to find inner and outer circle of a washer.

Labeling connections is explained in this article.

Solution (AVS)

  1. In Workspace Explorer open workspace Examples and in Film strip window select Washers dataset. Drag the Image channel to the ACQUIRE section.

  2. Add the ThresholdToRegion filter. Set the inMinValue to Auto and the inMaxValue to 200.0 to extract washers.

  3. Add the SplitRegionIntoBlobs filter to obtain separate washers.

  4. Use the RegionMassCenter filter to calculate washer centers. Click on the outMassCenter twice and change its name to MassCenters

  5. Add the CreateCoordinateSystemFromPoint filter and connect MassCenter to the inPoint.

  6. Add the FitCircleToStripe filter. Connect Image to the inImage. Connect the outCoordinateSystem to the inFittingFieldAlignment, that way you will only need to specify fitting field for one washer.

    • Run program to refresh the image and coordinate system, they will be later used to specify fitting field.

    • Specify the inFittingField. In fitting field editor select washer with coordinate system inside:

    • Set the inScanCount to 20 and the inMaxIncompleteness to 0. This will make the filter detect only complete washers and skip broken ones.

    • Using Show/Hide Ports show the outInnerCircle.Center and the outOuterCircle.Center. Change their names to InnerCircleCenter and OuterCircleCenter

  7. Add the PointToPointDistance filter to measure distance between the centers of inner and outer circle. Change the name of the output to the WasherErrorOrNil

  8. To create histogram of distances:

  9. Create a step macrofilter and name it DrawResults.

    • Connect the ReadFilmstrip output to the DrawResults to create a new macrofilter input. Check Array checkbox.

    • Connect the MassCenters and the WasherErrorOrNil to the DrawResults in order to create new inputs. Check Array checkbox.

    • Double-click on the DrawResults to edit the macrofilter.

    • Add a new formula to generate labels for washers. Create the Real? input inDistance.

    • Create formula outputs:

      • outString, type String?:

        (inDistance == Nil) ? "Invalid" : ((inDistance < 0.5) ? "OK" : "NOK")

      • outIsWasherOk, type Bool?:

        outString == "OK"

    This will mark washer as valid when distance between centers is less than 0.5 and the washer is not broken.

Macrofilter Main

Macrofilter DrawResults draws inspection results on input image

Used Filters

Icon Name Description
CreateCoordinateSystemFromPoint Most often used to define an object alignment from results of 1D Edge Detection or Blob Analysis.
CreateHistogram Creates the histogram of the array of real numbers.
RemoveNils This filter is useful e.g. when inspecting multiple objects while some of the inspections may fail (Nil result). RemoveNils is used to ignore the failed cases in the final result.
SplitRegionIntoBlobs Segmentation of a region into individual objects when the objects do not touch each other.
PointToPointDistance Measures the distance between two points.
RegionMassCenter Computes a point with coordinates equal to the average coordinates of the region's pixels.
FitCircleToStripe Precise detection of ring-shaped objects, whose rough location is known beforehand.
ThresholdToRegion Extraction of a region of objects that can be defined by a salient brightness.
DrawStrings_MultiColor Draws strings (text) on an image with multiple colors.

Further Readings