You are here: Start » Program Examples » Tablets

Tablets

Aim

The aim of the program is to detect missing or damaged tablets on a conveyor belt.

Input

A series of images from the conveyor of a production line.

Output

Detected missing or damaged tablets:

Hints

To detect damaged tablets it is recommended to consider using a DetectMultipleCircles filter together with a CheckPresence_Intensity filter to verify object presence by analyzing pixel intensities. To ensure that the inspection is performed once for an every batch of tablets, you can specify a Region Of Interest, where the inspection should be performed, as an input of CheckPresence_EdgeAmount filter.

Labeling connections is explained in this article.

Solution (AVS)

Add an EnumerateImages filter to get load consecutive images of some disk directory. Double click on the outImage and change its name to Image.
Add a CheckPresence_EdgeAmount filter.
- Change the name of outIsPresent to IsKeyFrame.
- Connect outImage from the previous filter to inImage of the current one.
- In Properties window in the bottom left corner set inEdgeScale to 2.
- Set inMinAmount to 0.27 to determine the minimum amount of the pixels that should be detected.
- Click on the CheckPresence_EdgeAmount filter and select inRoi parameter in the Properties window to mark the region as on image below:
Add formula which converts an input image into Nil when no tablets are found:

KeyFrame = if IsKeyFrame then Image else Nil
Add a DetectMultipleCircles filter.
- Connect input DetectMultipleCircles.inImage with Formula.KeyFrame.
- Set inRoi as on the image below:
- Set inRadius to 7 (approximate radius of tablets), inMaxOverlap to 0 (to avoid overlapping), inMinScore to 15 (in order not to miss any tablet) and inEdgeThreshold to 6 (the minimum accepted edge magnitude).
Add a CircleBoundingRectangle and connect its inCircle with DetectMultipleCircles's outCircles
Add a CheckPresence_Intensity filter.
- Connect Formula.KeyFrame with inImage.
- Connect CircleBoundingRectangle's outBoundingRectangle with CheckPresence_Intensity's inRoi.
- Set inMinContrast to 18 to avoid detecting bad tablets as good ones by checking their contrast.
Now create HMI by clicking on HMI Designer in the upper toolbar.
- Add View2dBox and connect outImage to its input inImage. It is going to display the images from the production line.
- Connect IsPresent with View2dBox.InStatus1.
- Connect TabletCircles with View2dBox.InData1.

Macrofilter Main

Used Filters

Icon	Name	Description
	EnumerateImages	Emulates image acquisition with images stored on disk.
	DetectMultipleCircles	Detection of circular or close-to-circular objects like holes, pins, pills, particles.
	CheckPresence_EdgeAmount	Quick and easy presence verification, e.g. for missing caps, screws, labels.
	CircleBoundingRectangle	Computes the smallest rectangle containing a circle.
	CheckPresence_Intensity	Quick and easy presence verification, e.g. for missing caps, screws, labels.
	DelayByPeriod	Suspends the program workflow for inTime milliseconds relative to the end of the filter's last invoke time.