Discussions

Low cost hyperspectral imager

Posted by RaptorUAS MD on September 24, 2014 at 6:57am

Hi,

 Interested to see if there would be interest for a low cost hyperspectral imager (N bands) lightweight sub 300gr with a spatial resolution of 2-8 Mpixel per spectral band.

 Projected unit cost would be $12,000 or less. We estimate to be able to produce 10 bands with a relatively narrow band at 20-50nm per spectral band.

 Thanks

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –
Follow

Replies

    • Darius Jack > Martin July 10, 2015 at 9:50am

      Martin,

      how much is your HSI sensor/ camera ?

  • Darius Jack July 8, 2015 at 4:11pm

    I have studied hyperspectral imager by NASA

    and employed hyperspectral imaging analysis to detect faulty, low-performing PV solar panels on Solar Farms world-wide.

    What can I suggest is to start from Light Hyperspectral Imager at $120

    then move to Medium Hyperspectral Imager at $1,200

    and finally move up and up with a price if market is already built and demand clear.

    Do you know any demand for Light Hyperspectral Imager at $120 ?

    darius 

    manta103g@gmail.com

    • Michal Rus > Darius Jack July 8, 2015 at 4:37pm

      Can you define "light" and "medium" ? AIS sensors used by NASA are surely not coming from these categories..

      I guess if it would be so easy, it would be done already but as RaptorUAS pointed out, you need a significant investment to create such low cost device and bring it to the market. That's not that easy..

      There are already lightweight AIS sensors available but it's still over 40000 USD. And by that I mean 100 narrow bands, not 10 wide bands system which falls into multispectral category..

      • Darius Jack > Michal Rus July 9, 2015 at 4:37pm

        Hi Michal,

        It took me 5 years to make a shift from Global Warming to Climate Change terminology among UN Agencies and other bodies.

        I have spent 2 years with multispectral imaginery of the Sun's activity provided by NASA to establish relations between Sun's activity cycles and Climate Change

        to build another better-grounded terminology for UN Agencies.

        I know market demand for multispectral imaginery and developed $100 spectroscope .

        All I need is 10-band multispectral imaginery to study Urban Heat Island,

        to study insolation level to develop solar farm for a given geolocation.

        Just let me know why do you need 100-narrow-band multispectral imaginery or hyperspectroscopy ?

        Following Vienna Conference I develop 100-DIY-seismograph network (DIY $100 seismograph for home use, to detect infrasounds transmitted through the ground, to detect vibrations from highway).

        So I am flexible in cutting development costs.

        I use drones to study solar farms for broken, faulty PV solar panels and

        multispectral imaginery in basic mode works fine.

        So I develop small r&d projects and my investment risk is contained

        since I am not $$$ NASA.

        If you can tell me why do you really need 100-narrow-band hyperspectral imaginery

        (medicine - cancer detection???) so I will try to find a nice technology solution for you.

        • Michal Rus > Darius Jack July 9, 2015 at 6:19pm

          Well why do you think there is a need for hyperspectral sensor ? Why are there already some at the market available ? I have a feeling you are trying to compare apples and pears. After my 3 years research on airborne hyperspectral imagery with those expensive sensors I think I can better understand now why it is needed. My application was aerial archaeology and although we did have only RGB and AIS data of the same flight available (so no multispectral to compare), I believe hyperspectral sensors and their special image analysis algorithms can provide better results. Imagine a GSD of 1m and 1cm but in a spectral way. You just see a much better results at the end. It all depends on what you are looking for but generally hypespectral narrow band sensor provides much more detailed information of any object. Similarly building a spectral signature from multispectral and hyperspectral data of the same object with the same defined spectral limits would bring vast difference. If you can build a cheap hyperspectral sensor, say 400-1000nm, 100bands, max 1kg weight including storage device, and do not ruin yourself in doing so, let me know.

  • Shaun Bell July 8, 2015 at 3:20pm

    Our company would definitely be interested.  We're looking for a light payload to work with a Phoenix lidar scanner.  

  • Nasruddin Abu Sari July 7, 2015 at 11:26am

    Yes, go ahead ! I am looking even for less than 10 bands system ( as long as it is better than current NGB / pseudo NVDI camera implementation)

    Price really matter for wide acceptance .. remember pseudo NVDI camera wins because of

    its price/performance offering !

    BTW where is the link for your "new Sateleye fixed wing"

  • Thomas Stanley-Jones July 3, 2015 at 10:33pm

    We would certainly be interested in a camera like that.  We need a camera to break into the hyperspectral realm, but it is too pricey at the moment, when we aren't sure what products we'll be able to produce for clients.  The multispectral camera's are too confined.  Good if you are only looking for one thing, but we aren't sure what we're looking for yet.  So yes, this sounds like a very interesting camera.

  • T3
    Thorsten July 3, 2015 at 2:16pm

    Hi RaptorUAS,

    Even though I think 12k$ is far too much to be attractive, I am pretty sure there is interest in such products - even with less than 10 bands. So please keep us updated about the development!

    Best regards,

    Thorsten

  • Sol Milne June 19, 2015 at 4:43am

    Hi, sounds interesting, what kind of spectral range does the camera have?

    Best wishes,

    Sol

This reply was deleted.

Activity