Extract height info from 3D Surface

[Guruthilak]

Hello,

I have an array of 2D data(which is the height information of disks taken from a sensor placed in XY plane).

Now i need to extract the Height information for all the disks and need to calculate other info from this height. Has anyone done this sort of exraction before or could help me in getting a solution. Please find the VI attached with the data samples

[altenbach]

"Array min&max" from the array pallette?

[Guruthilak]

No that wont help.. It will just give the min and max of the enitre array. I am looking in to height of each disk...

I opened up the block digram for the surface graph but was not able to figure out which vi wasactaully plotting the data. If i could find out probably i can extract the info? I am not sure just guessing...

[altenbach]

Then use a FOR loop and autoindex over each disk...

[Guruthilak]

Thats the real problem.. The disk are not aligned veritically or horizonatlly. Most of the time they overlap (as seen in the digram) and hence it becomes tricky to extract the data corresponding to each disk...

[altenbach]

Sorry, I was posting by phone. Cannot look at your VI.

You said you have a 2D array, where is the information kept that relates array element to disk number?

I will look at your VI later...

[Guruthilak]

There is no information that relates to disk number. The scanner/sensor would juxt scan any number of disks kept on the XY plane and gives out a 2D array of data.

I just plotted this 2D data on the surface graph, and the plot was almost accurate. So i believe the one fecade vi's whcih plotted out this information must have the complete information regarding the height data for each disk.. 

[GerdW]

Hi Guru,

the height data is in your 2D array.

You need to apply some filtering to your data:

- remove background by threshold comparison, that way you will know the location of your discs.

- find the center for those discs and use this to separate the discs. According to your example VI the discs are NOT overlapping, in the worst case they touch…

[Guruthilak]

I agree that the disks wont overlap ( what i meant to say that they are not alligned vertically or horizontally. My bad)

I was trying for the same as suggested by you. let me see how it goes.....

[GerdW]

Hi Guru,

you might start using an intensity graph instead of the 3d surface:

Here a simple threshold comparison gives a nice black/white image much easier to search for interesting spots…