GAProctor

There are now many powerful, lower cost LiDAR sensors available. The issue is putting them to use. It is no easy matter to effectively mount a sensor to a drone and sync it with an INS or IMU to get any kind of usable output. There are other issues as well such as number of returns, FOV and signal strength to consider. The good news is the good sensors are coming way down in price and that will benefit the drone mapping community eventually. What we need now is for the price of survey grade IMU's and INS components like the applanix to come down in cost.

Chuck, What do you mean by "really liked the developers"? I used Zephyr a couple of years ago. It was OK, though I did like it's ability to take frames from video, but I know they have been upgrading things pretty consistently.

What Do you mean by "Map Out"? Most generic cad programs have issues with large point clouds and so you need to find a program that is particularly designed to handle them. There are many of these and they each have differing capabilities. So part of the answer will depend on what you want to do with the point cloud. If you are modeling pipes and small detailed infrastructure, you want one that specializes in that. If what you want is to be able to run contours and do some simple structure outlines, then there are many that can handle this. What kinds of features are you trying to extract? Also important is what kind of underlying cad programs are you using as many of these are tied to one or the other. ACAD has a couple of different programs from simple to very complex. So does Bently. But there are many smaller stand alone programs that will do well and export to many file types like dwg or dxf. Things like PointCab or Pointfuse may suite your needs. Again, it depends a lot on what you are trying to accomplish, what kind of learning curve you are willing to take on, and what your budget is. Gary

Does anyone have any experience with using a Hyperspectral or Multispectral camera with a P4P (or other UAV) for mapping SAVs? How about lake/freshwater plants in general for use by a biologist to identify species and map locations? Or just ideas or notes in general?

First, I would say that time will tell with the P4-RTK. Let's see how it works in the real world and how various mapping scenarios effect it's performance. It is an advancement for the P4P in that DJI has provided a built in solution to the problem of timing the GPS data precisely to the picture. Most cameras have a hot shoe for the flash attachment that can be used to time the data, but no such thing is on the P4P, and that timing is critical and is also the focus of 3rd party RTK adapters for the P4P. The new P4-RTK does not change any of the previously noted differences between RTK and PPK. The nice thing with the P4-RTK is that you can use it both ways. I have never been in the PPK is obviously better camp. It is better if you are concerned about the issues that you have with possible loss of signal with RTK. But those can be mitigated and the advantages of RTK, in my opinion, out weigh the risks.

I was not aware that Aerotas did processing? What do they charge? how many photos? What kind of reports and deliverables? Can you do Nadir and Oblique in the same run? To me, that last is critical. Do yourself a favor and get a trial of P3dT and try it out. We are total Carlson, but even they told me probably not for what I wanted. I do have Carlson Point Cloud on top of Survey. I also do scanning and so use that for registering different clouds, but mainly so I can draft on the cloud and send it straight into my drawing. In the end, I went with Pix4D desktop. I will just go monthly ($350) as I need it, and work it into the cost of a job. You get all the tools you need to process and can draw nodes and polylines right on a tiff quality mesh. It is the only one that I found that works that way. If you have ever tried to pick around on all but the most dense point cloud, you know how troubling that can be. Also, download a copy of CloudCompare (freeware). It is an indispensable tool.

Great answer by Av8Chuck. My thoughts were pretty much the same in that it really depends on the size of the project and the desired accuracy. The Z axis is always the hardest to get consistent, repeatable measurements of, so knowing all the inherent limitations is very important. The thing with the GCP(s) is that it is more than an anchor for the Z measurement. They also are crucial in ensuring that warping of the image it kept to a minimum. So to me it seems like you will definitely need more than just the one GCP or you won't really be able to compare flight to flight.

Thanks Dave, I have to get more into both those apps as they are not as straight forward as some others and I didn't see right off where to do some of that. I see with DD that you have to add apps. I'm very happy to have more oblique options!

What are you using for your capture platform? IOS or Android? I have not had a problem using P4D capture on my IPad (though the other day it refused to upload the mission with my DJI P4P). I have several capture apps installed. Maps Made Easy, Drone Deploy, DatuflyPro and DJI GS Pro. Many people really like Litchi which I am going to try out soon. My main gripe about most of them is you cannot control the gimble. P4DCapture and for the some applications, GS Pro allow this (as I am told Litchi does as well), and only P4D has crosshatch grids which are nice for getting better detail.

Kind of off topic, but have you used P4D and Photoscan Pro (PSP) much so far for other projects? I have an older version of PSP and it has issues because it uses International foot instead of US Survey foot. I thought I had a fix for that, but my clouds keep coming in in off a few feet and have to be registered. It's easy to see, I just import the control points back into the dense point cloud and see if the match the GCP's. This happens even though I state the projection as NAD83 US WASouth survey foot. I hope that is fixed, because I do really like how well the clouds turn out. I am also curious how P4D compares to PSP. Have you done any head to head comparisons? Gary

Kermange, I'd have to wrap my head around this for a bit, but initially it seems like it should be somewhat straightforward. The "pit" as it were, is simply an upside down pile, so processing it should be no different than doing standard pile volumes. Maybe the software is having a problem with the "negative" direction, much like some older surveying software I had would go sideways if the coordinate values were negative. Gary

The problem here is that it seems you may have things a bit backwards. Also, without knowing how you controlled both data sets, it is almost impossible to give you a reliable answer. The first step should have been to ensure both sets of data are on the same system. You would need known points in both sets in order to do this. Using GPS derived points makes this easier, but for volumes it would not be necessary. GCP's are relatively easy to use in Photoscan Pro, but you need to know where they are, at least in relationship with each other, to be of any use. It may be easier to use the dense point cloud out of PSP and then bring the two data sets into any number of point cloud programs (CloudCompare is an excellent open source program) that will help you make the calculation. I will admit that I am not used to working with meshes, so maybe that can work. My strong suit is point clouds. The problem is that working with them can require specialized software beyond the normal cad package. CC for instance will allow you to "register" your clouds together. Just treat your cad points as any other point cloud, just like LIDAR. I have not done it, but I be Global Mapper with the LIDAR package has the tools to do it as well. But again, the data sets need to, some how some way, be on the same system horizontally and vertically. I wrestle with this all the time as I incorporate slam scan data into survey data (conventional and GPS) and LIDAR to create maps. I am just now starting a drone imagery division for my surveying company and hope that the lessons learned from terrestrial scanning will smooth out some of the bumps!

There may be others, but I no Datumate says their flight planner can do this. I have not used it but I believe they have a video that goes through the process. If you figure something out, please let me know.

I agree with @R Martin, but would add Photoscan Pro as an option as I personally liked the software better than Pix4D and actually think it has a much easier learning curve.

I too am about to purchase a drone to get my survey/mapping/aerial imagery business up and running. I am a licensed Land Surveyor (LS) in Western Washington and am looking forward to making the leap. Prior to being an LS, I was a licensed pilot and the UAV Coach Part 107 school to review for the exam and passed with flying colors within 2 days of the test being made available. After spending quite a long time researching at various models, including custom and self built, and taking the UAV Coach course on mapping, I have settled on the DJI Phantom 4 Pro. I likely will not do the plus as I want to use an IPad due to the greater flexibility in applications than when using DJI's tablet, even though it does have the very nice brightness. There are other more expensive drones and camera's, but while the 20MP, mechanical shutter camera that comes with the p4p is likely the lowest resolution I would go, it will do the job I need and the price point cannot be beat. Particularly enticing are all the collision avoidance and RTH function built into the p4p. A custom rig would be too many headaches to get started with so that was the first thing off the list. As for software, I would advise taking the course offered here for to see the flight control applications, and see the differences. Then just try each out. They are not particularly expensive. As for imagery processing software, if you are not going with something like Maps Made Easy that does the work for you, I would definitely recommend Photoscan Pro. I tried about every one of the various programs on a real world project and Photoscan was by far my choice. Pix4D is good as well but Photoscan's workflow and ability to process were my favorite. May of the other programs are variations of the same thing (which is some of French, I thin, program). That's just my opinion and I know Pix4D is popular and making changes and improvements all the time. It just didn't work as well for me. So there's my 2 cents! Gary