I am looking at buying/building a hexacopter for farm use. Looking to use a multispectral camera on it (like the parrot sequoia, agrobotix or agrocam).
I need at least a 20min flight time. Payload is not that heavy. Those cameras together with gimbal should not weigh more than 1kg.
I was looking to build one based on the Tarot frames.
While researching I found this built which pretty much matches what I thought would be fit for my purpose:
Does anyone have any experience with this type of built for farm use?
Or any experience with any frame for farm use?
Maybe you can suggest a better solution than the one I found?
Any input is much appreciated.
Hi, I like share my mapping experience with RPAs or drones, first you don´t need gimbal, the program you use could correct the minor camera angles. Second, any multicopter is not suitable for long photogrametry areas. You must use a fixed wing dron. Third but not least, you need a dedicated camera to obtain a good information, in this aspec, inexpensive cameras are not good invertion. modified cameras only capture IR light, but you need at least NIR and and red o RGB to obtain NDVI indicator.
You can see this example link.
Is not a NDVI map, but is the same work with a different camera.
Good luck and best regards.
PS: English is not my native language.
I think a Phantom 4 pro is not a flexible enough platform.
And not suitable for agricultural imaging. Also a quad.
Dji M100 or Inspire1 seem better but they come at quite a price.
Gary Mortimer said:
The less overlap and no cloud processing features are quite attractive.
Probably a custom gimbal can't be that expensive. Surprised there isn't one that would fit the camera. Maybe in a few months.
Aglyx seems to be exactly the app for crop management via ndvi.
andrew cutter said:
andrew cutter said:
Whats wrong with a Phantom 4 Pro?
I hadn't seen the AgroCam units before, probably since it is EU based. They look like modified GoPro H3 clones. It's a good price for a starter NDVI camera...similar to my modified S100 with the Event38 NGB filter. The prices on ag mapping can get crazy so if you're not doing it professionally, it's best to start with lower cost components. As Andrew mentioned, the higher priced cameras improve sensitivity but may be pricey for entry level NDVI.
I lost my subscription to Pix4D so I am now trying the Fiji ImageJ app with Ned Horning's plug-ins for NDVI. The FireFLY6 Pro model is rather expensive but it is a solid performer with APM Copter under the hood. My DIY FireFLY6 was much less expensive but you need to know what you're doing to make it work.
The VTOL concept, along with all the take-off and landing benefits, opened the door for greater mapping area coverage in a given time period. The FF6 Pro can cover 450 acres (182hectares) in a single 45 minute flight so my DIY FireFLY6 can cover half of that. This concept is now being tested on the relatively new QuadPlane features of APM Plane. My Ranger EX hasn't flown a mapping survey yet but it can hold multiple cameras and can fly longer than the FlireFLY6 for a fraction of the cost.
You'll find alot of the NDVI results are subjective, so, if you start with lower cost cameras, you can customize your own baselines for your specific farm and crops. Here are a few videos from my initial survey flights.
FireFLY6 Mapping Test 1 from Gregory Covey on Vimeo.
FireFLY6 Mapping Survey 2 from Gregory Covey on Vimeo.
Slant range and red edge are price about the same . You get higher resolution , less over lap required (which means more acre covered and more images ) , on board processing and the ability to analyze the picture with your laptop vs going to the cloud with red edge to have it process and analyze. You save processing cost vs the red edge . one of the problems with slant range for me is mounting it to the quad copter . I would have to have a custom made gimbal (current their isn't one available ) . I have a farmer friend fields to learn on . wheat , soybean and corn . Its important to have a few season under you belt . The ability to compare different years helps you to know how good of job you are doing or not . Store the images in the cloud is a must . nvdi images take up alot of memory. Crop management . - helps with inputs (roi) , plant stress , pests , cuts down ground truthing or crop scouting , predict what type of harvest you might have , helps with crop damage from storm (nvdi helps with getting the real damage cause to your crops , not what your eyes can only see ) .
keeping track of all the software for analysis is quite a challenge. they are also intertwined like you say. this industry is dynamic.
agribotix has some demonstrations of drone deploy use with a phantom 3 and their camera solutions.
was not convinced by the phantom 3 as I need to cover 50Ha initally in as little time as possible. but Drone deploy looked viable for a trial. thanks for your suggestion.
so slantrange does it with less than 60% overlap? thats quite helpful of course.
I am not sure how it compares to the sequoia and the rededge. it is priced in between at $4k.
i was thinking to use this as a starter: http://www.agrocam.eu/cameras. maybe its better than it seems.
aglytix looks good. slantview looks good but at a price. like you said, it advertises only 20% overlap.
so you are using a spyder 6 frame with rededge that you rent and process on dronedeploy?
what area do you survey? and do you have a few seasons worth of data on the same land plots? would be interesting to know if doing this whole imaging thing actually helps with crop management.
andrew cutter said:
thank you, good info.
the FireFLY6 looks great, a little expensive for me to start with.
the Ranger EX Quad conversion looks really good. Can you fit both S100 cameras on the Ranger?
Also what area are you able to photograph with your planes and MRC?
I am looking to cover 50Ha (btw i am based in Europe) to start with. Ideally it should not take more than 3 hours of field work.
The s100s look like a nice good value solution for ndvi. I also found http://www.agrocam.eu/ . good prices.
Do you use pix4 software to process the images? I found that the software for processing/analysis can be quite expensive.
Greg Covey said:
A modified camera won't give you the accuracy or quality that you really want . Good place to start to learn on or if you are on a really strict budget . But i wouldn't bet the farm on it or use it to make decision on . The chip they use is not the best design for nvdi for agriculture . Their is no way around this right now . I've seen the prices drop quite a bit over the last two years of the camera you really need to use . My understanding that the software companies that offer subscription based system , allow you to rent them for short period of time ( the way i go until prices drop alot ) . As far as size of copter i would go 750 or large for line of sight . I seen people paint them orange to make them more visible . Sky hero allow to you modified the copter to a larger one . So if you go with their 650 size you can always enlarged it later. I've even seen as of the last two week more ag analytic software companies coming on line. So i'm guessing prices will be dropping soon .