Would Mavic be suitable for photogrammetric purposes?

[PeteM]

I see your point Pete, but the specs aren't that far apart between the P3/P4 and the Mavic so the new drone could fill it quite alright even though it hasn't been tested yet. I for my part am looking for PORTABILITY because that's a real constraint at times, and the Mavic deliver in spades.

Portability is not a significant factor for me.

The Mavic strikes me for all redundant aids and sensors that engineers have implemented into it. I think the Mavic will be safer and easier to "drive" than the Phantom series... but it will do the work for which I buy it ?

Another concern I got is the resistance in "difficult scenarios" like taking off and landing on hard terrains... Panthom has a structure that separates sensitive parts of the ground, the Mavic has gimbal and camera almost grazing the ground level...

 

[AlexSP]

Yep, sort of GPC made easier. Did you get a chance to try it?

 

[r62ewa]

I haven't had the opportunity, just talking with the company and getting specs. The major drawback I see is they are given a 2 year lifespan at present.

 

[AlexSP]

Portability is not a significant factor for me.

The Mavic strikes me for all redundant aids and sensors that engineers have implemented into it. I think the Mavic will be safer and easier to "drive" than the Phantom series... but it will do the work for which I buy it ?

Only you can answer that

Another concern I got is the resistance in "difficult scenarios" like taking off and landing on hard terrains... Panthom has a structure that separates sensitive parts of the ground, the Mavic has gimbal and camera almost grazing the ground level...

I see that as an improvement. The Phantom needs the "legs" because it has a large, clumsy (IMO) gimbal whereas the Mavic has a really neat, small and "tucked" camera. Now, not only is it less exposed, it also helps with GC and stability of the aircraft during all times, above all during take off and landing but also during flight.

Phantom legs are no big help if the AC suffer an impact. I see that much better on the Mavic, as the folding legs are coiled and seem to double as a "dampers" - though it isn't clear by how much (and I wouldn't rely on it anyway), the whole design looks more stable and protected overall than the P3 IMO.

Problems with debris getting in motors and gimbal? I can say for myself, but I take off and land with my DJI quads from grass and mildly-dirt areas, and never had a single issue with dust or anything. I avoid as best I can, and some places are no-no of course, but doesn't really bother me much. If it gets something on it, I'll clean it up, dust off and pack the bird. Sure the Mavic seem more exposed to this for sitting lower, but 1) we can always take off and hand-catch; 2) the gimbal has a protective bubble.

Either way, we'll have to wait and see.

 

[AlexSP]

I haven't had the opportunity, just talking with the company and getting specs. The major drawback I see is they are given a 2 year lifespan at present.

Cool, it looks promising but I'd like to have a field test or see one before adopting it. But it looks nice indeed. Will get in touch to get more info, good idea. As I said, if I look from the business perspective the rational is different, that's to say it might be worth it if the demand can pay for the 2-yr lifetime and if it is as precise and accurate as they tout.

RTK is the definitive answer, but - again - it has to justify the investment, because it's not a toy. And the whole stitching/processing for that level of accuracy is also more expensive too, must take that into account. So I'm waiting on some contracts that are on the pipeline to go for it.

Besides, technology is improving FAST in this niche, unless I can make some money out of something I find it's better to wait. For now I'm dealing with GCP but honestly most of my jobs don't demand 1 or 2cm accuracy so it's fine. I really like the P3P for mapping and modeling, let's see how the Mavic performs at this.

 

[PeteM]

The protective bubble must be removed during the flight... however, it can be solved with a takeoff/landing area built for the purpose...

Yes... I think that as soon as the first models will begin to arrive in the homes of those who bought them the answers will start coming... and we'll also see how app developers will act...

 

[Deanster]

OP, I'm just baffled by your posting.

Mavic has different optics than P3/P4.

If they're not well-suited for your specific needs, then get something that works for you, or if Mavic has other features that are valuable, figure out how to make it work with Mavic's optics.

Your notion that somehow Mavic or DJI is 'wrong' or 'lower' is just loony.

All products have compromises - GoPro's wacky fish-eye, DSLR-toting big drones, etc... all have size/weight/flight time/optics trade-offs built in.

Being PO'd that this particular set doesn't meet your needs seems a bit extreme, especially when it sounds like you can just adjust altitude and have a very similar set of imagery.

There's no reason 40m is required or somehow special, it's just how high you had to fly a P3/4 to get a decent set of imagery for your uses. Treating it like it's a sacred altitude that can't be violated is... weird.

 

[tenpine]

I haven't read all the replies on this thread but I'd like to point out that professional photogrammetry from standard aircraft is rarely done at low altitudes with wide angle lenses. Flying higher is generally better because it minimizes scale distortion when you are flying over uneven ground or 3D objects.

 

[43k]

I feel I can say that this drone is ABSOLUTELY not suitable for this type of application.

It's actually the opposite.

The longer the focal distance (a narrower field of view) that you use for this kind of work, the better the images will work in any software designed to stitch images together.
Especially when the ground is not perfectly flat, or where there are trees or buildings for example.

As others pointed out here, with a longer focal length you get the exact same resolution using a higher altitude. Objects on the ground map to exactly the same number of pixel in the image.
So that's not a con either.

Also, a longer focal length normally gives you less distortion, which is perfect for this kind of job.
So the Mavic is actually more suitable for photogrammetry than the P4 in this respect.

 

[Bobby Jones]

Don't give up just yet, actual testing will commence soon. I use a P4 for forensic studies in conjunction with terrestrial scanning. The lower flights give better results simply because of the increased detail and clarity. I too see a potential issue with the Mavic Pro. Simply increasing the altitude is not the answer. While more images at lower altitudes is certainly an option, but more images require more processing time and results in larger files, both of which I want to reduce. I look forward to trying the Mavic Pro for these studies and will report back soon.

Information not insults make a great forum.

 

[tenpine]

I don't understand the "increased detail and clarity" aspect of the argument. That would only pertain if there was a great deal of atmospheric dust, fog, etc., but in most real world use this would be insignificant. Or if the lens was inferior but usually the more a lens runs toward telephoto, the better the resolving quality, especially in the corners.

 

[David_Cambridge]

The other issue to consider is barrel distortion from both lenses. Very wide angle lenses have large amounts of rectilinear distortion.The Phantom 3 lenses are pretty good, but have noticible barrel distortion towards the edge of the frame. You can see it easily using Lightroom, which has built in DJI lens profile compensation. Switching it on and off will show how much distortion there is.

The Mavic, with it's longer focal length lens, should have less of it. I would think that better rectilinear performance would be much better for photo metrics

 

[Jody Pitts]

I simulated (through a professional software) various photogrammetric flights with Mavic.

I feel I can say that this drone is ABSOLUTELY not suitable for this type of application.

Setting a flying height compatible with the accuracy required to the ground, setting just a discrete overlapping and speed parameters, due to the reduced FOV, the complexity of the flights (number of strips and photos) increases in an unacceptable way.

The issue ends here for me... the optics of this drone is not suitable for making topography...

I hope that many will throw away their P4 for Mavic so I'll be able to buy my first Phantom at a low price...

Take this post down so I can get one as well bro!!!

 

[pelermon]

Go to 10% overlap

 

[skiptv]

I simulated (through a professional software) various photogrammetric flights with Mavic.

I feel I can say that this drone is ABSOLUTELY not suitable for this type of application.

Setting a flying height compatible with the accuracy required to the ground, setting just a discrete overlapping and speed parameters, due to the reduced FOV, the complexity of the flights (number of strips and photos) increases in an unacceptable way.

The issue ends here for me... the optics of this drone is not suitable for making topography...

I hope that many will throw away their P4 for Mavic so I'll be able to buy my first Phantom at a low price...

Totally great post!!

 

[pierce]

To a great number of users, camera it's not all that matters.
You surely get a phantom 4 at a low price as soon as mavic begin populating in the drone comunity.

 

[With The Birds]

Portability is not a significant factor for me.

The Mavic strikes me for all redundant aids and sensors that engineers have implemented into it. I think the Mavic will be safer and easier to "drive" than the Phantom series... but it will do the work for which I buy it ?

Another concern I got is the resistance in "difficult scenarios" like taking off and landing on hard terrains... Panthom has a structure that separates sensitive parts of the ground, the Mavic has gimbal and camera almost grazing the ground level...

Pete it is surprising how little clearance is afforded by the phantom landing gear between the camera and ground. Uneven ground, sand, grass etc you really need to carry a small landing pad to comfortably deploy in a lot of situations. Do this for the Mavic and you have an added advantage, it can auto land within an inch of its take of point using vision positioning. If you had a complete controller failure the Mavic would be able to land safely on return, the phantom may be 6 or more feet off. An 8th inch price of plywood cut 4 inches larger than the area described by the landing gear should be perfect (with a simple pattern applied to the face to make life really easy for the VPS).

 

[metronome]

perhaps if money is no problemo, try something like this....

Automated Drone System | Kespry

 

[Ohary]

I think Pete ought to revise this post as his testing is simulated and not actual. I also tend to agree with the others that have stated a narrower angle lens should result in less distortion effects. The Mavic should in theory produce better results (supposed softer images, not considered).

I also believe it comes down to the amount of data. I believe you compensate with height above target, and/or ovarlap. I believe there will be a solution if this does present a challenge, but coming to conclusions without actual testing is misleading.

 

[Nick Mendocino]

Be careful about posting that anything is ABSOLUTELY true. Especially when you haven't even done any real testing. A little knowledge is a dangerous thing.