Lens selection for close proximity surface photos

AI Thread Summary
The discussion centers on finding suitable lens options for capturing images of surfaces just 5mm from a CMOS camera sensor. Suggestions include using a lens with a very short focal length, employing mirrors to extend the object distance, or considering a pinhole camera design. Participants highlight the challenges of distortion due to the extreme angle of view required and recommend exploring alternatives like a "door peeper" lens for compactness. The conversation also touches on the importance of illumination and optical performance, suggesting that reflective optics might mitigate some issues. Overall, the design challenge is acknowledged as complex, with various innovative solutions proposed.
Danmo
Messages
4
Reaction score
0
Good day all,

I am searching for some advice on how I could successfully take photos of a surface that is 5mm away from a CMOS camera sensor.

My ideas so far are;
1.) Find a lens with sub millimeter focal length (0.2mm by my calculations)
2.) Utilize mirrors to increase object distance
3.) Pin hole camera design

I have attached a sketch to illustrate.
 

Attachments

  • object distance.png
    object distance.png
    4.5 KB · Views: 220
Science news on Phys.org
Why not increase the object distance and use a longer lens?
And welcome to PF, by the way.
 
  • Like
Likes Danmo
The vertical space allowed for the camera is approx. 15mm.

Oh, and thanks!
 
Danmo said:
The vertical space allowed for the camera is approx. 15mm.

Oh, and thanks!
If you have a space to one side then you could use a 45 degree mirror and a long lens on your sensor. But then, I'm not sure about the usable field there would be. I guess the mirror would need to be curved (convex) and placed at the far side of the subject.

Only you know the actual geometry but I wonder what sort of distortion the arrangement in the diagram will produce - and what masking of the surface there could be, if it's not flat all over. Could there also be a question of available light and depth of focus, whatever system you choose?

A few more sketches are required.

A
 
  • Like
Likes Danmo
Danmo said:
Summary:: Lens selection for short object distance

Good day all,

I am searching for some advice on how I could successfully take photos of a surface that is 5mm away from a CMOS camera sensor.

I have attached a sketch to illustrate.

This is going to be a very difficult design challenge, if your diagram is reasonably accurate- a 100mm diameter object placed 5 mm away subtends nearly 86 degrees (170 degree total field of view), so you are likely looking at a fisheye design with all the distortion that implies.

What about a 'door peeper' lens like this:

https://www.grainger.com/product/1H...ucid=N:N:PS:Paid:GGL:CSM-2295:4P7A1P:20501231

Not sure it will fit your space.

That said, you haven't specified some basic performance aspects: I assume this is visible light imaging, for example. Are there any optical performance requirements regarding throughput or resolution? What about the object illumination- is this a transmission or a reflected light situation?
 
  • Like
Likes Danmo
Thank you both sophiecentaur and Andy Resnick for your input.

Great idea with the door peeper Andy, I will look into this!
And yes, the object will be illuminated by led lighting therefore is a reflected light situation.
 
I wonder if the fisheye distortion can be (effectively) removed in software.,
 
I believe you need to use a macro lens and a macro flash, it is best to use a tripod at the same time.
 
I think your idea #2 is the best bet. Use a mirror to relay the image to where-ever the camera fits.

If you have never looked at the workings of a flatbed scanner, find a cheap or junked one and do so. Closest to the object surface there is a flat mirror. From there the path is a series of curved mirrors in a folded light path (3 or 4 mirrors if I recall correctly). The advantage of reflective optics is they don't have to be massive for short focal lengths, also no chromatic aberrations.

Of course a scanner uses a line sensor which is much shorter than the paper width, so the mirrors are curved in only one dimension. It sounds like you need area sensing though, so two-dimensional optics are needed. Anyhow, the optical path is light weight and compact -- just not as compact as you are after.

Is this for a production item or can one (a few) be sacrificed for one-off data collection?

Please keep us updated on you progress.

Cheers,
Tom
 
  • #10
Mobile phone cameras are extremely thin and have increased in there ability to produce good results - May be a way forward
 
  • Informative
Likes Tom.G
  • #11
Perfidium said:
Mobile phone cameras are extremely thin and have increased in there ability to produce good results - May be a way forward
It would make a lot of sense if it were possible. I wonder if the lens may be too integrated for a one-off modification, tho’. The spec details would be important. It might be interesting.
 
  • Like
Likes Perfidium
  • #12
Or this type of thing for ideas
lots of this sort of thing about in various configs
eBay item number:
353209858759
 

Similar threads

Replies
4
Views
2K
Replies
22
Views
5K
Replies
14
Views
3K
Replies
3
Views
2K
Replies
152
Views
10K
2
Replies
52
Views
7K
Back
Top