What You Are Saying
Readers share thoughts on surveying trends, more
POB wants to hear from you! Let us know what you think about any POB content in print or online by sending an email to Editor Perry A. Trunick at firstname.lastname@example.org. You can also send a Tweet to @POBMag, leave a message on the Point of Beginning Facebook page, connect with us on LinkedIn, or visit the RPLS Network at www.rpls.com.
Re: Drone-Based Photogrammetry
Thanks for bringing this subject up. It is the first time I see someone questioning it. As a photogrammetrist and professional land surveyor creating high-accuracy planimetric and topographic maps for many years, I posted a conversation in the GIS-Photogrammetry Group site related to the same question and have not gotten any reply/answers yet.
UAV seems to be the “Silver Bullet.” Nowadays we hear and see it everywhere. It is being presented as the miraculous fix for everything and a cut through all complexity, providing an immediate solution to problems. Land surveyors are very excited with it; “We can do photogrammetry now.” Then the propaganda goes further, saying: “…photogrammetry from acquisition to CAD faster and at lower cost.” Is it?
For years we have done photogrammetry with large-format cameras and regular aircrafts, and we have been very careful about it to ensure quality product is delivered that can immediately be used by our clients — either as a map in the old years or as a digital CAD or GIS file in more modern ages. For the last 10 years, the creation of orthophotos and surface models with contours generation using manned aircrafts and large-format cameras has been pretty automated (UAV workflow is just taking advantage of the same technology). And it does not take too long to do automated surface for a 2- or 500-acre site (sites where UAV would be efficient for acquisition of data).
I think it is important to consider what products are needed when receiving a UAV dataset. Or are you looking just for a pretty picture, with a cool 3D view of your site with a cloud of millions of colorized points for you to interpret/delineate? Do we deliver it to our client and let them deal with it? I see UAV capable of collecting hundreds more images to perform the same tasks, requiring a lot more control points to guarantee engineering accuracy. Sometimes the site does not allow collection of all needed control points if you can’t access it. Also, when it is time to produce a surface model that can reliably be used as a ground surface by the engineer working on the site, it will, just as in the large-frame format camera, require some intensive editing to remove bushes, parked cars, fire hydrants, etc. and identify hard breaks that require breakline.
I am curious to know which software or means are being used to perform the same “photogrammetric products” as the drone-based photogrammetry folks claim, with all planimetric features collected, and if anyone has insight about whether or not it is really more cost efficient, considering the number of images, the need to access the site, the additional control points and the fact that the camera is not a calibrated camera.
– Luiz Cortes