Comments

• 

  David March 20, 2013 at 10:59pm

  LanMark- If your primary interest is developing image processing software, I think the MaxMax cameras are the perfect camera to trying things out with. My primary interest is streamlining the production of othromosaics so whole fields can be mapped quickly in an automated fashion. The less I have to worry about the spectral validity of a camera the better (at least in my situation). PixelWrench2, the software application that comes with the Tetracam cameras can output a variety of vegetation indices including NDVI. Typically I create the CIR orthomosaic, then I import that back into pixelwrench to perform the NDVI calc and apply a color pallet. Below shows how a "healthy" and "stressed" sugarbeet leaf. You can compare the RGB, vs CIR and NDVI (the CIR came from the ADC Lite and the NDVI was derived form the CIR using PixelWrench2)

  

  

  
• 

  LanMark March 20, 2013 at 10:32pm

  Martin.. completely agree.. did you see my posting today on the drone crashes.. they are 30-300% more likely to crash than a similar sized airframe... article here
  
  So yeah.. unlike what Chris says.. they are NOT uncrashable and they are NOT safe... drones are flying lawn mowers that are often heavy and could kill someone if it falls on their head.
  
  There is a long way to go before that is possible... or that anyone would feel safe with that.  However we have drone farm tractors and other ground based equipment already... not sure if they have sprayers.... one would think the farmer could just hit 'go' on the sprayer and have it go to town.
• 

  LanMark March 20, 2013 at 10:26pm

  David.. I am not sold on anything.. well outside of buying a MaxMax camera to play with.. and as a software dev, I have a tendency to tinker and code.. especially if I don't like what the company provides for imaging processing.  However.. I would be interested in whatever works... $3800 might be a bit high for some precision farmer... but like most things.. it if has results then it can be expensed as this is all a business.
  
  I have a tendency to think that with some sort of calibration target/card that you capture from the altitude you are flying and the camera settings that you are using..  I would think that there would be some way to baseline/calibrate the maxmax camera data that you captured.. which you likely should/would need to do with the Tetracam one, since a lot has to do with time of day, and other environmental factors of the day of your flight.  I haven't explored if this is even possible but seems like it could be.. since the calibration card would have known NIR, R, G, B elements.
  
  I would love to take a look at some of that Tetracam TIF outputs and what algorithms people use as they would be different from the NIR-G-B cameras... is that something you could share with me?  especially if you also have the NDVI colorized output from whatever GIS app.  I would love to bake in the ability to use more then just MaxMax cameras to do this sort of analysis... MaxMax just happened to be at a price point I could play with.
• Martin Davies-Roundhill March 20, 2013 at 10:23pm

  Hi, I was responding to the following post on the forum.

  Comment by Troy Reabe yesterday

  I think the field of avation needs alot of advancments before a drone can be able to start to replacecrop dusters like the Thrush or AirTractor. There are helicopter drones but the have a very low coverage rate per hour. Also Helicopters are expensive to run per hour.

  For our spraying service during the high time of the year for us there is not enugh time in the day to do the work that needs to get done and we are able to do around 200 acres per hour with a plane provided we have a short ferry time. A helicopter like the Bell 47 can't even do half that.

  We are also carrying 500 Gallons a load. so this would be a big drone or you would need a nurce truck on sight.That has its own problems of a portable chemical containment system for mixing and loading the drone.

  Cheers Martin

  drone australia
• 

  LanMark March 20, 2013 at 10:16pm

  Martin.. there has been no mention of crop dusters in this forum.
• Martin Davies-Roundhill March 20, 2013 at 10:00pm

  Hi, Troy Reabe, Your comments are very thoughtful and you are certainly knowledgeable. 

  In your comments,

  "I think the field of avation needs alot of advancments before a drone can be able to start to replace crop dusters like the Thrush or AirTractor. There are helicopter drones but the have a very low coverage rate per hour. Also Helicopters are expensive to run per hour.

  For our spraying service during the high time of the year for us there is not enugh time in the day to do the work that needs to get done and we are able to do around 200 acres per hour with a plane provided we have a short ferry time. A helicopter like the Bell 47 can't even do half that.

  We are also carrying 500 Gallons a load. so this would be a big drone or you would need a nurce truck on sight.That has its own problems of a portable chemical containment system for mixing and loading the drone.""

  The above assumes that drone equipment will be large or a t least "our" size.

  I would suggest to you Troy that many drone activities will be conducted by very small craft, think bee size , and like bees there will thousands who will work together literally in a swarm. 

  Looking at the crop dusting task from this angle avoids many of your concerns and it is exactly this simplification that is a key strength of drone technology.

  I continue by reffering to your comments on crop assessment for disease control using drones.

  The same drone system swarm that crop dusts would also be used to monitor disease either . With 2 huge benefits. Disease can be detected at a very fine level, to the plant or even part of plant, and treatment can be made by same drones . Clearly this will mean huge savings in treatment materials and crop lossess. It is to be imagined that over time a farm with drone swarms would have  nearly zero disease and require very little consumable resources to do so.

  In such a drone swarm system the farmer only has to maintain supply of materials in storage "hives" which the drone can access autonomously.

  The farmer has to task the drones but once done its almost set and forget and the drones will just proceed with very little assistance.

  Troy I hope you find the above food for thought and look forward to your comments Cheers
• 

  David March 20, 2013 at 9:39pm

  LanMark, you bring up some very excellent points.

  -The Tetracam ADC cameras output RAW format images that contain reflectance information for the Green, Red, and Near-Infrared bands. By not performing onboard jpeg compression, like I am guessing the MaxMax does, you are preserving a more true representation of the canopy reflectance. The software included with the Tetracam cameras uses a color processing file specific to each camera (made from a calibration standard) to convert the RAW images into Color Infrared TIFs, or whatever standard format you would like.

  -I personally have not used MaxMax cameras but they are definitely a cheaper alternative that I would be interested in learning more about. I do get a little nervous flying a $3800 camera around, but they are very robust since they have no moving parts (i have heard of people really crashing the Tetracam cameras hard with no damage).

  -For 90% of ag applications, a 3.2 megapixel camera flying from 400-1000ft agl will produce higher resolution data than can be effectively used. Wide angle lenses are bad for remote sensing applications...the ADC cameras use 8mm lenses giving a 44 by 34 deg field of view. 

  -For Directed Scouting applications (aka the imagery is only being used to indicate where potential problems might be) the actual pixel or NDVI values are not so critical. I would however be hesitant to derive a prescription map from a non-remote sensing grade camera systems since the pixel values are going to affect a chemical rate applied. Also the benefit of having good absolute NDVI values is that you can compare imagery taken on different days throughout the growing season. 

  I am very impressed by the software application you wrote, it looks like it could bring a lot of additional value to the MaxMax cameras
• 

  LanMark March 20, 2013 at 9:30pm

  Eric I friend requested you.. as I would love to get some more insights on the consumer side of it all..   the maxmax.com cameras are in the $700 price range.. I got the modified Canon SX230HS.. which has its own gps.. I figure even if the GPS wasn't all that great I could at least get great timestamps on the imagery without having to sync the camera's clock.
  
  Anyway I will be more chatty privately as I tend to ramble on.. but I think people like you are looking for something they can do that isn't all that expensive and provides an alternative to walking the crops.. but allows them some decent way to detect problem areas to investigate in person and with other equipment.. 
  
  Do you guys do precision farming?  There are all sorts of fun stuff you can do with GPS-enabled equipment..  I have no idea how to generate the data that those things consume.... yet.
• Eric T March 20, 2013 at 9:24pm

  LanMark,

  I'm new to the UAV scene but am currently building a quad with APM and my brother is working on a plane with autopilot as well.  We should have a decent setup for taking some images of our crops this year to start to get in on this.

  We farm just over 3000 acres, mostly irrigated, in Alberta, Canada.  We aren't 100% sure how much use we'll get out of aerial images but are into the UAV hobby anyways, and are always looking for ways to improve our day jobs (farming).

  Approx what price range are you talking for a MaxMax.com camera capable of taking the types of images you are working with on your custom MaxMax software?  And will you be needing testers for your software?  Or at the very least we could possibly send you some images in the future of our actual fields to play with if you need more source files.

  I agree with your statement about this not being an exact science.  Most farmers have a good routine for crop production including walking the fields, and aren't going to be replacing their current systems anytime soon, but more information is never a bad thing.  Most farmers also generally have a good idea of the problem areas in their fields, but having more detailed and documented information is also a good idea.  

  A trending tool is exactly what we need.  Eventually it would be nice to have an overlay of our entire farm for every year on GoogleEarth or something so we could easily monitor poor areas over time.  I guess eventually more than once a year would be great, but it would take some time to cover our entire operation more than once a year....

  Anyways, very interested to get involved in the conversation!

  Eric
• 

  LanMark March 20, 2013 at 8:42pm

  Awesome stuff Keeyen... love the airfoil you guys have on your site from Germany.