I would like to create a high res map of an island (Koh Mak in Thailand). The island 4.5 x 3.6 miles. The highest point of the (overall pretty flat island) is 300 feets. See here
I am pretty lost in the huge amount of information available so I ask the forum:
Do you think it is possible to create an image of these 5.5 square miles?
Can the job be done in a team (a freind has the same DJI Mavicd 2 Pro) and then combined in one photo?
The end result will also be a poster in better quality than what can be extracted from GoogleEarth.
PS: I just calculated that with the FOV of the DJI Mavic 2 Pro Camera being 77 degrees I would need aproximately 82 photos at 500 m height (with no overlap though). My PC should be able to handle that.
Of course it can be done, but I’m not sure how you are going to get away with flying at 500m. I assume you got that from the DJI recommendation of maximum height? I looked up the drone laws in Thailand and it states the maximum altitude is 90m AGL. Beyond that there are also VLOS laws. So unless you have some extreme waivers in order to abide by those laws it would take you approximately 35 batteries and some 12,000 images which you would not be able to house in one map on DroneDeploy, but that could be flown as several maps and then combined and/or served in GIS software.
Hello. I know flying at 500 meter is not allowed. But I have a special OK from local authority and I am in touch with the only plane that ever flys in this area. So when this happens I will be “official”. Will DroneDeploy be able to house 905 Photos?
I wonder how flying at 500m would be like with a Mavick2. A lot of battery will be used to get to that altitude and to return, and then how well will the remote transmit out to that distance and elevation.
Seriously, no overlap? Even at that height your going to want to have some overlap, albeit you won’t need as much. Especially with the rolling shutter on the Mac 2…but then I haven’t seen this done before at this scale.
Extreme drone mapping. Can’t wait to hear the results!
No no. I do not mean I will do without overlap. That was just my mathematical calculation on how many pics would be needed. It will take some time. I have to learn first. but I will publish the result.
Who said no overlaps? Did Iiss it? Personally I wouldn’t use any less than 55 even if all I wanted was a 2D map to look at. You really need to have 3 consecutive images overlapping if you expect any accuracy in measurements. Especially when there so much homogeneous terrain. Worst thing I can think of is flying an effort like this and it not stitching. Second worst would be getting a map that you can’t do anything but look at.
You cannot fly at 500m that is in aircraft traffic area. An 1-5cm accurate Orthomosiac Map could be generated with several dozen flight missions it might take a lot of flying, the higher the better, within the transport guidelines for your area probably 400 feet. I did a small section of an Island comprising 26 acres and it took more than 3 hours of flying. Hopefully that helps calculate your work. A 60-70% overlap is necessary for a good outcome. I also created an Oblique 3D model though with takes 5 times the images from a variety of angles to create the model. Getting ready to 3D print it.
That’s allot of flight time for that area. My planning software tells me that I can fly 25 acres at 400ft with 80/65 overlaps in 7 minutes. Excluding home to waypoint and back to home time which is usually less than 5 minutes.
Factor in the pauses of 2.5 seconds for each shot as we both are shooting with a Mavic 2 Pro that does not have a mechanical shutter so you can’t shoot on the wing without stopping like the Phantom 4 Pro.
The reason for me to be looking into this project is solely for creating a good map of the island to be used on the web and on print.
Right now though, I will not start with the project due to very bad air quality which would make the photos look bad. Like every year agricultural waste is being burned all over Thailand and especially over Cambodia and we sometimes have air quality at the level of downtown Bangkok. Additionally we are still in what normally is “high season”. No rain since months means dry land not looking good.
So I will wait for the beginning of the rain season. Rain will wash out the air and vegetation will look more inviting. The map should also be publicity for Kho Mak
Just to update the readers of this topic: I start to realize how complicated the project is.
I lowered the flight height from 500 meter to 200 meters. That increased the amount of images from 1080 to 6660 and the flight time from 150 minutes to 339 minutes. That means 22 batteries. In an optimal timing that makes 6 flights per day (I only have 3 batteries). So I will be flying (as long as I am alone) for an esthimated 4 to 5 days. Beside the different light situation between morning and afternoon I will also have the trouble of different light from days to day.
Since my only goal is a map (that can also be printed in large scale) that will generate a final picture with different exposure/light in different pictures (as taken day by day).
I could not find any reference data to explain this problem. I doubt people fly missions that take days…
I am reluctant to decrease the reslution (currenly at 4.5 cm / pixel). It wouldn’t reduce my flight time anyway.
I also do not know what will happen with the area of the island that has high “mountains”. Most of the island is about 0 to 150 feets over the sea. One section though has what the locals call mountains. These are around 300 meter high. Since terrain awareness is not available for plans over 200 acres (I am slightly above that with 2300 acres) I have no idea how that will affect the photo quality.
People fly 1,000’s of acres and multiple days all the time, just not usually with a quadcopter or with DroneDeploy. I am sure that there are many that fly fixed wing with other software that process with DroneDeploy, but the majority of the users here probably don’t fall in those categories. You can however alleviate much of the exposure variation by post-processing your images in something like Photoshop. We choose to use ACDSee. You can even do some color matching if you feel it necessary. Obviously you will want to fly at the same time each day, but if you need to space it out according to weather it would be best to fly on all sunny or all overcast days if possible. Don’t mix and match or the resulting map will be spotty.
300 meters or 300 feet? If you plan to fly at 200 meters AGL then I hope it’s feet… …and if that’s the case by flying that high you shouldn’t have any issues with stitching a 2D image. What overlaps are you planning to use?