GoPro Hero 2 camera

I'm curious if anyone has tried using one of these small video/still cameras on a microscope. It is super wide angle and when I held mine up to the eyepiece for a test, it recorded the image as a central circular spot (about 60% of the sensor). It got the whole view through the lens, but I would prefer capturing the image using the entire sensor. Thought perhaps I could make a stack from the HD video to use in depth enhancing software.

Henry Barwood
Troy University
Troy, Alabama USA

Hi Henry,

That's an interesting little camera. I haven't seen it before, but something tells me that I have a use for one. I'm just not sure what that use is yet..

If you have a 2X tele extender laying around, you might try it in front of the camera lens. That will narrow the entrance acceptance angle such that the sensor will be more fully illuminated.

I think that you would have to use the single shot mode to do multi-focus. A video could be taken as the specimen is moved in the z-axis, but you wouldn't have any control over the step sizes or number of steps. Maybe you have some other idea?

Gene

Hi Gene,

I actually tried a 2X Barlow, but the lens system on the GoPro is and ultrawide, almost fisheye, lens and it didn't couple well. I really need an ultrawide field telescope eyepiece, but those suckers only come in 2 inch sizes and start at the cheap end at around $400. I'm still working on the problem.

You don't have to control the step size to stack in video. Astrophotographers take videos and then convert them to a stack using something like Iris or ImageJ. You can then set the program to pick out specific frames to depth stack (or collapse like they do for astronomical imaging). All you would have to do is have a stage that raises the specimen while you are making the video. The software will do the rest.

Henry Barwood
Troy University
Troy, Alabama USA

Hi Henry,

There is a bit of misinformation in what I previously said regarding the tele extender. I mistakenly thought that the GoPro lens was removable and that a tele extender could be kludged in somehow. I also said “in front of”, which should have said behind. Oh well, another senior moment.

I'm not too surprised that a Barlow lens doesn't couple, as it is a small lens and most likely cannot be mounted physically close to the camera lens. I solved the problem of matching my Sony F-717 to my microscope by building a coupling lens from two 50mm diameter 80mm fl achromatic lenses in a symmetrical design. I agree that a wide field telescope eyepiece might work. One of these days, I will have to try one of my TeleVue eyepieces borrowed from one of my other hobbies.

I understand what you are saying regarding stacking from video. Some issues that I can see with this approach are:

If you raise the specimen too fast, the increments will not have overlapping focus. If you raise it too slow, then you get more images than you need. In this case, sorting out the images that just overlap focus may be difficult. If the stack is unnecessaraly large, the processing time can become an important issue. I tried processing a stack of 10 each 8MP images, and it took 20 minutes. That's not entirely unreasonable, but my goal is to process a stack of 100 each 18MP images. You can see that as the stack size increases, the processing time tends to become asymptotic to never. I am now having to get a new computer just for that purpose, but even then, processing time will be an issue. I was impatient waiting for film to be developed and now I'm impatient waiting for stacks to process. Perhaps that speaks to my age....

I don't mean to imply that the video method is not viable, but it comes with its own set of issues to overcome. If there were no challenge, there would be no fun, so it is certainly worth a try.

Gene

Hi Gene,

With the video stacking programs, you can scan through the stack and select images for a sub-stack. A scan of a specimen raised on a stage might take 5 seconds at 24 FPS. That would give you 120 frames. A quick scan and you could select 5-10 of them to process in the Z axis. Many DSLR cameras have even higher resolution video capture, at slower frame rates. When I have a bit of time I'll play with it and let you know how it works.

Henry Barwood
Troy University
Troy, Alabama USA