Machine Vision

cancel
Showing results for 
Search instead for 
Did you mean: 

Machine Vision fundas (Circle Detection)

Circle detection is very interesting function which is used for detection circles in binary image to put it simply but this function has lot of application. There is no direct function in labview to determine the width of say an irregular pipe. This function I find it very handy and useful.

 

In the forum I have answered lot of question on how to extract the circle detected and how to overlay the circle found.

 

ROI from circle detection

 

The circle detection does not return a clear bounding box for us to overlay or extract. The "circle data" return the centre x,y and the radius.

 

18091i1AAA15172132630C . The rectangle top,left and bottom,right can be found as shown below

 

 

x-radius  - Top

y-radius  - Left

 

x+radius  - Bottom

y+radius  - Right.

 

With this using a build array we can extract or overlay the circle found. But If we want to extract multiple circles, We can do the following

 

18093i58B70553BA19926F

 

 

A for loop as shown above can be used to extract individual circles. Again the vi shown above will display only the last circle extracted. There are lot of way you can use this to change the way you view the result.

 

Example

 

Processed image                                                                       Input image                                                                         

 

  18099i657BAAC5D57C767B18095i3C1770FED1DFCD46

 

Overlayed image

 

18103i9B1A733EC98DA424

 

 

 

 

 

Message 1 of 4
(5,353 Views)

Muks

I fopunf a way to only extract the circle and not its surroundings. You could use the image

mask and the  work to get only the circular region and not the square. Have you tried it.

 

 

yeshwanth

Message 2 of 4
(5,232 Views)
Hello, How to get the circle without its surroundings. Could you share your idea or Vi? thanks. Attached is the processed pic.
0 Kudos
Message 3 of 4
(4,641 Views)

attach processed image,thank you

0 Kudos
Message 4 of 4
(4,640 Views)