Switch to English Language Passer en langue française Omschakelen naar Nederlandse Taal Wechseln Sie zu deutschen Sprache Passa alla lingua italiana
Members: 75,777   Posts: 1,671,092   Online: 977
      
Results 1 to 2 of 2
  1. #1
    Mustafa Umut Sarac's Avatar
    Join Date
    Oct 2006
    Location
    İstanbul - Türkiye
    Shooter
    35mm RF
    Posts
    4,300
    Images
    110

    Multispectral Photography

    I have a scientific gear catalog and there are lots of cheap holographic elements for diffracting the incoming light to analyze. I know telescope people uses the same with extreme expensive ccds.

    If I make a slit panoramic camera like Noblex , how could be possible to place hologram between film and slit ? How can I calculate the distances ?

    And can I extract from blurred image, an non blurred image after scanning and for example a copper intensity map inside a mountain photograph ? How this is being done ?

    I think if this is done stereo and load in to augmented reality , would be very exciting.

    Umut

  2. #2
    Mustafa Umut Sarac's Avatar
    Join Date
    Oct 2006
    Location
    İstanbul - Türkiye
    Shooter
    35mm RF
    Posts
    4,300
    Images
    110
    Here is the answer from astronomy com by Jeff B.

    OK, I understand now.
    You're going to be some distance from the target rocks and will be "sweeping" a surface area and recording the light reflected back at you.
    There are a number of issues you'll be facing, and I confess I do not know them all. But here are some ...
    Even with a high-resolution spectrograph, it's unlikely you will be able to identify specific rocks. Let's say, for example, you're imaging a talus slide. You would be able to get reflections from areas of the slide, but those areas most likely will be larger than the individual rocks. If what you're doing is a general "survey" then this will be okay. But if what you want to do is to be able to look at an image, then go back later and pick up a specific rock sample, I don't think you'll have the resolution to do that unless your spectrograph to sample distance is on the order of a few meters at most. If it's tens of meters, the individual reflections are going to merge.
    To get a high-resolution spectrogram, you are going to need a fine grating, and that will "eat" a lot of the light -- so daylight or strobe are probably your only options. If you're imaging something like a rocky mountain slope, then you're left with daylight. If you use sensitive film and daylight, plus you make your panoramic scan slowly, you might be able to gather enough light to illuminate a broad slit (but that leaves you with lower resolution). So I think you're going to need a large aperture (to gather more light) and that generally comes at the cost of field of view (to get a good image of a distant target with high resolution through a slit normally means a longer focal-length, hence, less field of view. A spectrograph requires a specific (or very small range) of focal lengths to reach a good focus. You could get a decent focus and exposure with a 50 to 150 micron slit using a cooled CCD imager or pushed film with an F10 catadioptric telescope, but this wouldn't be a very portable system unless you went with an apeture on the order of 6 inches or less -- again, this would mean either a short spectrograph to target distance or a longer exposure to get enough light on the slit. With a distant target, you have much less control over the lighting (incident sunlight, for example).
    Your spectrophotometer or spectrograph is likely to require a good power plant. I use a cooled-CCD spectrograph setup (for stellar spectrography) that requires at least one laptop and a decent-sized 12V power plant to handle the current requirements. And the longer you image, the greater the power requirement. The larger the powerplant, the bulkier and heavier it will be. This might lead you to some sort of tradeoff.
    The spectrograph will require some sort of collimating lens to ensure the light rays hitting the grating are parallel. The reflected or transmitted spectrum will then need to be captured with a "taking lens" inside the spectrograph. For an area survey, you're going to want something more like a spectrograph and less like a spectrophotomer. If you were evaluating hand or lab samples, a spectrophotometer would work better, allowing you to collect spectra from specific crystals within a sample, for example. So, for an "area survey" the spectrograph ("camera") will require a focusing lens inside ... this forms the focused image at the film plane.
    I can envision two kinds of images: an area image, much like a landscape photograph, but with spectra accompanying the image highlights (specular reflections), or a "rainbow" image, where the spectrum is a long, narrow rectangle with each "bar" in the rectangle being an image of the slit (illuminated by the light from the target), where the slit is continuously scanned across the scene to form the spectrum. Each "bar" or column location in the image will correspond to an image of the slit over a specific part of the scene. In the first case, correlating parts of the image with a specific rock or area of the landscape will be easy, but the resolution will be determined solely by the resolution of the grating and the image scale. In the second case, the resolution will be a combination of slit size and image scale, with the grating resolution another factor. The grating resolution may be higher than the slit width, but if so the slit width will be the determining factor. It will be difficult to correlate slit position with area of the landscape, so you will need a way to move the slit and grating out of the light path without moving the spectrograph (you must not change the aiming point) and make another "normal" panoramic image. You could then compare the two images to find the area that corresponds to a given slit position in the spectrogram.
    I'm not aware of surveys that have been done this way, but it certainly seems possible. The remote-sensing spectrographs I do know about are designed to work at "taking distances" of thousands of kilometers (think a spacecraft orbiting Mars, or the Moon), so even if the instrument's resolution is high the effective resolution at the target's surface will be measured in kilometers, or at best hundreds of meters.
    One way to test this before coming up with a final design is to obtain a mounted grating, place it in a cardboard tube, and visually scan a typical target area in bright daylight, to see how the image changes as you move the grating (scan the scene). You can obtain mounted gratings and telescope-eyepiece adapters from companies like Rainbow Optics and Baader Planetarium. You could make a simple camera-mounted transmission grating spectrograph and see how well it all works. And then you could move on from there to a panoramic camera and / or full-blown spectrograph.
    It sounds like an interesting project.



 

APUG PARTNERS EQUALLY FUNDING OUR COMMUNITY:



Contact Us  |  Support Us!  |  Advertise  |  Site Terms  |  Archive  —   Search  |  Mobile Device Access  |  RSS  |  Facebook  |  Linkedin