How much ground can I see?

[Tenly]

Good point about the x and y axes. It's been a few decades!

Your extrapolations using the ratio areas make sense to me, as the rectangle of the proportions has to fit within the circumference of the circle of the FOV. Now go out and fly, to confirm it!

But @WithTheBirds has suggested that your calculations might have determined the hypotenuse instead of the x dimension as I supposed.

If the FOV is stated as 94 degrees, does that take the sensor into account or only the lens? If only the lens, how do we know how much of that falls upon the sensor.

I also had been picturing the field of view as a rectangle which is why i thought your calculations would refer to the widest dimension - but hearing you call it a circle reminded me the lens itself is round so I think @WithTheBirds is probably right that the FOV is the diagonal (hypotenuse) and we need to do some extra math to get length and width.

C^2 = 2.144 * height
A^2 + B^2 = C^2

For a 4:3 ratio sensor, we have a 3,4,5 triangle - so to make things easy, to get 50 for our hypotenuse, our height would be:
50 / 2.144 = 23.3208m

Therefore at 23.32m altitude, the captured area on the ground would be 40m x 30m!

Did I screw that up?

 

[GadgetGuy]

You may well be right. My suspicion was based on the fact that a 20mm focal length lens mounted on a traditional full frame camera (3:2 aspect ratio) produces a field of view of just over 94deg (longest corner to corner dimension).

I think it is the other way around. The 20° equivalent was arrived at by finding a full frame focal length which has the same 94° FOV of the P3 camera, which as you stated is a 20mm lens. I don't think the aspect ratios of each are used in the FOV calculation, but, clearly, they have an impact upon the results.

 

[GadgetGuy]

But @WithTheBirds has suggested that your calculations might have determined the hypotenuse instead of the x dimension as I supposed.

If the FOV is stated as 94 degrees, does that take the sensor into account or only the lens? If only the lens, how do we know how much of that falls upon the sensor.

I also had been picturing the field of view as a rectangle which is why i thought your calculations would refer to the widest dimension - but hearing you call it a circle reminded me the lens itself is round so I think @WithTheBirds is probably right that the FOV is the diagonal (hypotenuse) and we need to do some extra math to get length and width.

C^2 = 2.144 * height
A^2 + B^2 = C^2

For a 4:3 ratio sensor, we have a 3,4,5 triangle - so to make things easy, to get 50 for our hypotenuse, our height would be:
50 / 2.144 = 23.3208m

Therefore at 23.32m altitude, the captured area on the ground would be 40m x 30m!

Did I screw that up?

You guys are doing great! The rectangular aspect ratio is inscribed within the circle, and the calculated diameter of the circle (2.144 times the height) would, in fact, be the hypotenuse of the right angle formed by the two aspect ratio sides, so it would be a 3,4,5 triangle, and I calculated the hypotenuse side.

Your final calculations make sense to me!

At a height of 23.32m, you capture a 40m x 30m rectangle on the ground, centered at the GPS coordinates of the aircraft, where the diagonal is 50m.
Just plot out how many rectangles you need for coverage, allowing for roughly 25% overlap, or whatever you think you can get away with for stitching purposes, and start shooting. There are also expensive commercial apps that will automate this whole process for you. MapsMadeEasy.com is one of them. You simply enter in the coverage area, your type of aircraft to provide the FOV and the aspect ratio, and it creates an autonomous mission that the drone will fly to provide full 3D mapping of the area. The 3D rendering software from the images is the expensive part, and is usually licensed by the month. For 2D, you might be able to use their flying app for free, and just stitch in Photoshop or LR.

 

[Tenly]

There are also expensive commercial apps that will automate this whole process for you. MapsMadeEasy.com is one of them. You simply enter in the coverage area, your type of aircraft to provide the FOV and the aspect ratio, and it creates an autonomous mission that the drone will fly to provide full 3D mapping of the area. The 3D rendering software from the images is the expensive part, and is usually licensed by the month. For 2D, you might be able to use their flying app for free, and just stitch in Photoshop or LR.

Yeah - I tested out DroneDeploy with their 30-day free trial and created this 3D interactive map of aboit 20% of a nearby island that contains about 300 cottages. This was my first attempt with the software - and I know (now) it could have come out much better if I had flown orbits around the individual cottages - but as it is - the technology they use to do this just blows me away. This interactive 3D map was created in 5.5 hours using 401 still photos all taken at an altitude of 60m with 70% overlap.

1464370570_JIMOPENPIPELINE by DroneDeploy - 3D model

...and a zoomable, Google Maps style high-resolution 2D map.... It's cool how far you can zoom in.

DroneDeploy - Put your drone to work!

I have no professional use for this technology, but I would like to do a few more of these for fun. Unfortunately - the companies that I've seen that offer this are crazy expensive. DroneDeploy just had a sale - a license was either 5500/month or 5500/year - I know - big difference - but then again it's not. Both prices are way out of my budget for "fun".

Do you know of any companies that offer this technology for free or very cheap for non-commercial use?

 

[matti]

I made a LibreOffice spreadsheet of these calculations (I saved it also in Excel .xlsx format). Below is a screenshot.

Just enter the AC Distance (or altitude -- i.e. the distance from the P3P camera to the object) in meters, feet etc (the results use the same unit).

The results show the object's maximum diagonal diameter and the object's maximum X- and Y-dimensions for 4:3 and 16:9 aspect ratios.

(Edit: error in 16:9 calculations corrected 19.2.2017)

 

[GadgetGuy]

I made a LibreOffice spreadsheet of these calculations (I saved it also in Excel .xlsx format). Below is a screenshot.

Just enter the AC Distance (or altitude -- i.e. the distance from the P3P camera to the object) in meters, feet etc (the results use the same unit) and the aspect ratio (4:3 or 16:9 etc).

The result shows the object's maximum diagonal diameter (assuming the whole sensor is used with no cropping) and the object's maximum X- and Y-dimensions for that aspect ratio.

Please correct me if there are errors.

View attachment 57398

Haven't checked the math, but awesome!

 

[npalen]

Is there a misconception that the P3P camera takes better pictures than the P3A? My understanding is that they are both 12MP (4000x3000). Is there something I'm missing that makes the P3P a better "still" camera? Just asking.

 

[npalen]

Yeah - I tested out DroneDeploy with their 30-day free trial and created this 3D interactive map of aboit 20% of a nearby island that contains about 300 cottages. This was my first attempt with the software - and I know (now) it could have come out much better if I had flown orbits around the individual cottages - but as it is - the technology they use to do this just blows me away. This interactive 3D map was created in 5.5 hours using 401 still photos all taken at an altitude of 60m with 70% overlap.

1464370570_JIMOPENPIPELINE by DroneDeploy - 3D model

...and a zoomable, Google Maps style high-resolution 2D map.... It's cool how far you can zoom in.

DroneDeploy - Put your drone to work!

I have no professional use for this technology, but I would like to do a few more of these for fun. Unfortunately - the companies that I've seen that offer this are crazy expensive. DroneDeploy just had a sale - a license was either 5500/month or 5500/year - I know - big difference - but then again it's not. Both prices are way out of my budget for "fun".

Do you know of any companies that offer this technology for free or very cheap for non-commercial use?

Have you checked out MapsMadeEasy.com? Seems like reasonable rates and the processing is done with their servers.
For flying the grid, apps such as Altizure and Pix4d seem to work fairly well.

 

[kennedye]

Someone needs to go out to a large empty field, put down some colored chalk or something, and start taking measurements!

 

[Tenly]

Have you checked out MapsMadeEasy.com? Seems like reasonable rates and the processing is done with their servers.
For flying the grid, apps such as Altizure and Pix4d seem to work fairly well.

DroneDeploy can also be used for free to capture the images and their processing is done on their servers also. I have not looked at MapsMadeEasy, not Altizure but I will! Pix4D has a capture and a mapping app with free trials. I tried that but it runs locally on a PC. I have a super powerful iMac with 48GB of memory, fast quad-core CPU, solid state drives and an upgraded video card and it still took 8+ hours to generate the 3D map which didn't seem to look as good as the one I got from drone deploy. It was my first attempt and it was quite likely something I did wrong.

Thanks for the tips. I hope your idea of "reasonable rates" is similar to. mine!

 

[Tenly]

Someone needs to go out to a large empty field, put down some colored chalk or something, and start taking measurements!

I was thinking an empty parking lot might be a good place to check since it already has markings on the pavement that can aid with the measurements....

Thanks for volunteering! Let us know what you find!

 

[npalen]

Or a football, rugby, soccer etc field?

 

[Meta4]

Is there a misconception that the P3P camera takes better pictures than the P3A? My understanding is that they are both 12MP (4000x3000). Is there something I'm missing that makes the P3P a better "still" camera? Just asking.

They are identical cameras except for a chip to process the 4K video.
If you are shooting stills, there is no difference.

 

[kennedye]

I was thinking an empty parking lot might be a good place to check since it already has markings on the pavement that can aid with the measurements....

Thanks for volunteering! Let us know what you find!

ha, that's what I get for replying. I do know of an area down in Woodinville where I might be able to put out some distance markings without disturbing anyone, so if I get a chance this week I'll see what I can do.

 

[Tenly]

ha, that's what I get for replying. I do know of an area down in Woodinville where I might be able to put out some distance markings without disturbing anyone, so if I get a chance this week I'll see what I can do.

Ha ha! Thanks. I'll try too and then we can compare notes.

 

[GadgetGuy]

Is there a misconception that the P3P camera takes better pictures than the P3A? My understanding is that they are both 12MP (4000x3000). Is there something I'm missing that makes the P3P a better "still" camera? Just asking.

If you ever intend to install the HDMI out board to use FPV goggles, the bit rate on the P3A will sorely disappoint. It's an unadvertised difference that bit a few P3A owners, after installing the HDMI board for goggle use. The P3P HDMI output is significantly better quality. Also, the still jpg images extracted from the 24 fps 4K video on the P3P are very usable at 4-7 MB each! You get 24 of them every second! Try that with the still camera capture! However, for dedicated still captures alone, there is no difference. The price difference is also less now than at original release, over a year ago. It's about $100-$200, instead of $279.

 

[With The Birds]

It seems DJI claim of FOV being 94 deg for 35mm was accurate. It is even wider in the 4:3 sensor in the P3- closer to.100deg approx.

My rough measurements determined the angular corner to corner FOV for P3P is a 35mm full frame SLR equivalent of 96 deg which is just over 100 degrees with the 4:3 aspect ratio of the PSP canera.

I got the camera as close as i could to square on a masonry wall at 2m distance and placed tape at the extend to the visible area. I then measured distance between tape and to approx sensor position from wall and did the basic trig.

I had hoped to go back further than 2m however it was difficult to see where to.place the tape as everything was so small in the image.

 

[Tenly]

It seems DJI claim of FOV being 94 deg for 35mm was accurate. It is even wider in the 4:3 sensor in the P3- closer to.100deg approx.

My rough measurements determined the angular corner to corner FOV for P3P is a 35mm full frame SLR equivalent of 96 deg which is just over 100 degrees with the 4:3 aspect ratio of the PSP canera.

I got the camera as close as i could to square on a masonry wall at 2m distance and placed tape at the extend to the visible area. I then measured distance between tape and to approx sensor position from wall and did the basic trig.

I had hoped to go back further than 2m however it was difficult to see where to.place the tape as everything was so small in the image.

What if you just went ahead and placed two intersecting strips - 1 horizontal and 1 vertical - on the wall - as long as possible - and then snapped pictures of it at various differences? Couldn't you then zoom into the picture and read the numbers off the tape so you could do the math?

If someone had access to a large room with a floor using standard 12" tiles, it might be easy to estimate area also... The best I have access to is an empty parking lot or if try to do it myself...

 

[With The Birds]

What if you just went ahead and placed two intersecting strips - 1 horizontal and 1 vertical - on the wall - as long as possible - and then snapped pictures of it at various differences? Couldn't you then zoom into the picture and read the numbers off the tape so you could do the math?

If someone had access to a large room with a floor using standard 12" tiles, it might be easy to estimate area also... The best I have access to is an empty parking lot or if try to do it myself...

Tenly if that parking lot you have access to has masonry block walls it would be more suitable for this exersize,

I understand my rough test setup was not close to ideal for a precise measurement however arriving at a calculated angle of 82 deg accross the frame (which depicts a frame corner to corner derived angle of 96 deg for 3:2 frame format) gave me comfort that DJI's advertised field of view of 94 deg, also stated as 20mm focal length on traditional 35mm format, is likely accurate.

It also means the actual corner to corner FOV for the phantom3 3:2 frame, while not provided to us in the DJI literature, is in fact greater than 94 deg and likely closer to 98 deg.

 

[Tenly]

Tenly if that parking lot you have access to has masonry block walls it would be more suitable for this exersize,

I understand my rough test setup was not close to ideal for a precise measurement however arriving at a calculated angle of 82 deg accross the frame (which depicts a frame corner to corner derived angle of 96 deg for 3:2 frame format) gave me comfort that DJI's advertised field of view of 94 deg, also stated as 20mm focal length on traditional 35mm format, is likely accurate.

It also means the actual corner to corner FOV for the phantom3 3:2 frame, while not provided to us in the DJI literature, is in fact greater than 94 deg and likely closer to 98 deg.

I gotta admit I don't understand much of what you just said - especially the 3:2 frame reference. Is that the same as the "3:2 pulldown" I've seen referenced in some video editors? Is there an easy way to explain what it actually means? I can't be the only one that doesn't know its significance....