At Smart Mobility World 2015, our Italian representative, Andrea Marella, was interviewed by IOT Expo. We did not manage to get rights for modification of the video, so we bring it here in the original Italian without subtitles and add an English transcript below. (The auto-transcript and auto-translate feature from Youtube is not really worth trying out in this case).
(ENAC is the Italian Civil Aviation Authority)
|Journalist||Andrea Marella, from DataFromSky. I would like like to better understand in which field do work and who you are.|
|Andrea Marella||Thanks for the question. My job has been dealing with traffic and transport ambit for some years now. I discovered around one and a half year ago this service, that we are currently able to provide to our customers also here in Italy. Practically, we can record aerial videos, using drones or similar technologies, and by means of that we can analyze the video using our processing software. This allows us to extract much more information than we usually have, and not simply the static ones, but also dynamic data. So for example considering a given vehicle that has been recorded in the video, it is possible to tag it and see its movement, as well as know, for each video frame, the speed, acceleration, deceleration, time to enter, time to exit,… All these data are something that we, as traffic engineers, didn’t have until short time ago, but they’re available. Who could be interested in those data? Public administrations, traffic research centers, motorway companies, but also shopping malls. For example, we recently conducted some surveys around the Zara-Expo area in Milan, outside the Expo exhibition, where we had to evaluate and analyze the long-term flows. In conclusion, it allows collection of a large amount of data.|
|Journalist||So what you are talking about is an applicable solution, that has to do with the drone sector, and that could be very useful to compute data analysis. About the application sectors, are they related only to traffic analysis, or may they be different?|
|Andrea Marella||We developed this algorithm that allow to capture the information on a generic moving object. In this case we apply it on vehicles driving along a road, but it could be applied also, for example, to vehicles being parked, knowing the standing time of each vehicle, the enter and exit time, other different important vehicle parameters. As I told before, focusing on shopping centers, we also could estimate the queue, study possible gridlock conditions, the waiting exit time… And last but not least, the road safety, where many applications may use the data that we offer – because what in fact we do is provide information and elaborate them for the customer. This may be, as I said before, public administration, as well as an individual user.|
|Journalist||About the drone sector: It’s a field that is currently fast expanding, but at the same time faces lots of complication to freely grow, because of the problem related to the ENAC regulations, the need for license, authorization, restrictions,… At the moment work is a bit complicated. What do you think about it as an ‘inside worker’?|
|Andrea Marella||Let me start by saying that we actually provide the analysis service, and the survey phase is generally conducted by our partner companies, they have everything needed to legally and safely carry out work. What I can personally say, is that within the last 18 months we had two revisions of the regulations about flying objects; every time that I meet experts or other workers who have something to do with it, we tell each other: ”Next time will be the right one for changes!” I hope in the next time!|
|Journalist||I know. But is there a place in the world where it would be easier to start developing and launching on the market, US for example, thanks to the advantageous circumstances? In other words, is that a problem of rules, or related to how the regulations are applied in a given place?|
|Andrea Marella||I think that the regulation imposed by ENAC, that is one of the first ones adopted in Europe, still have space to correct and improve. Anyway they’re becoming aware of the problem, giving openness. For example the opportunity to fly with UAV less than 300g at 100m height on urban areas is a great possibility. On the other hand, we have UAV producers and workers, who have already invested resources and sums of money, and they’re ready to enter the market. There’s still lack of something due to, well you know in Italy, bureaucracy. For example, the need to send declaration to ENAC, ask authorizations… All these require time, days or months… more likely months! Next January probably a new application on the ENAC website will be born, that will be very helpful to simplify the procedures. I trust in this opportunity!|
|Journalist||From your point of view, what you see in the next two years? Because, as you know, the UAV sector has lots of practical uses… I’d like to know your thoughts – which applications are going to have an immediate development? And what is going to happen within the next two years?|
|Andrea Marella||Generally speaking about drones, I think that monitoring applications related to agriculture will be the biggest, and the first to grow, due to ease of work in open spaces, without constraints, safely. Next, there are research centers dealing with the general environment monitoring; I mean infrastructure monitoring, land monitoring, or related to wind farms, or, as we do, in the field of traffic monitoring.|
|Journalist||So an uncertain future, not well defined…|
|Andrea Marella||No, I wouldn’t say so. We’re optimistic, we don’t want to be negative. We believe that 2016 will be a crucial year for the real growth, and that we are going to see the number of flying drones increasing. Once we are able to fly them easily, we’ll have to decide how to optimize for usefulness: then we will be ready to take advantage of it!|
We worked on another demo video from Brazil, from Certare engenharia e consultoria. They perform a lot of turning movement and origin-destination counting. The camera was placed on a Phantom 3 professional from DJI, which was positioned perfectly over the area. It was one of the numerous junctions in Fortaleza, a junction of two one-way streets.
Traffic flow at the site is 2136 vehicles/hour.
If you follow our work for a longer time, you have probably noticed that so far we did not provide any exact numbers on accuracy and errors of the measurements. However, a method providing quantitative data without any quantification of its precision is somewhat dubious. Thus it should come as no surprise that we were working on these matters zealously behind the scene.
So, what do we have now? We have created a model of the whole process which generates our data, and we are currently working on a set of measurements to validate it.
We have created a mathematical model of the whole process which generates our data. This includes the physical reality at target, complete optical system, and digital processing. The following error sources were considered:
- landmark location errors (in meters)
- landmark pixel uncertainty (in pixels)
- camera intrinsic parameters (in pixels)
- target pixel uncertainty (in pixels)
- air turbulence (in pixels – included in target/landmark pixel uncertainty)
Using this model, we were able to relate together many variables of the setup – achieved accuracy, distance, covered area, incidence angle, slant range… The results were encouraging. The following picture shows area covered when using a 4k camera, depending on incidence angle and slant range, assuming maximal error of 0.5 meters:
What can one read from the chart? As you can see, the incidence angle of about 40° is a reasonable cutoff value. Slant range of 140 meters at 0° (i.e.: directly overhead) gives the best value. For a HD camera, the area covered is a quarter of that for 4k, and optimal altitude in zenith is halved – 70 meters.
We can also overlay the model’s predicted accuracy onto real pictures – that is, display achieved accuracy along with the footage. We hope to eventually incorporate that functionality into DataFromSky Viewer, so that you could check yourself. For now, we have this picture from the Randers video (in HD). Numbers are error in meters, with respective isolines displayed. A 4k video would yield half the error.
In order to validate the model, we made a set of measurements at a suitable place near Popice, a small Southern Moravian village known by the vineyards in the area.
We placed a regular grid of 64 landmarks in an 8×8 square pattern, so that a side of the square was exactly 100 meters. The landmarks were positioned using a professional GPS in differential mode, achieving placement accuracy of about 5cm.
Then, we set up an UAV to fly around and take a video, in 4k of course. Here is the trajectory projected onto ground, looking at the area from the west.
We simply imported the recorded video into DataFromSky and added the landmarks as tracked objects. You can’t see them in the picture because the red ID label “pin heads” are larger than an A4 at that resolution, but they are there.
We are still working on processing the results. So far, the agreement between model and measurements is very good, and the model output suggests accuracy greater than we hoped for!
We will publish the results in an academic journal paper. Hopefully, the paper will be finished in a few days and we will be able to share more!
Since this text is about accuracy, we can hint that there is more to come: We measured the vehicle position using a vehicle-mounted dGPS as well, so there is be another set of data to work on.Read more
Another video from Denmark analyzed, this time from a 4-way junction in the center of Charlottenlund and processed as a demo for WeFly. In slightly over 3:45, almost 140 vehicles pass, at an average speed of 10 kph. The traffic flow appears to be 2163 vehicles/hour.
There is a public school nearby, so there are a lot of student pedestrians in addition to cars.
The footage was taken on 20.11.2015 from the altitude of 57 meters. The camera used was GoPro 4 on a gimbal, recording in full HD at 50 FPS.
A few days ago we mentioned the boxed version of our software. Design of its envelope is based on our print brochure, which we will give away at events, conferences, fairs, and whenever the opportunity arises. And you can now browse this brochure online to take a look!
It might be better to mention that the brochure is not exactly small, so it might take a while to load after you enter.
If you are a partner, feel free to grab a print-quality PDF version from the partner download zone.Read more