We are proud to announce, that we become a member of NVIDIA Inception Program! This virtual accelerator program led by NVIDIA nurtures dedicated and exceptional startup companies, who are revolutionizing industries with advances in AI and data science, and help them in all stages of product development, prototyping, and deployment. Using NVIDIA Pascal GPUs and Caffe’s deep learning framework, DataFromSky is able to do the automatic extraction of time-space-trajectories of moving objects in aerial video data captured by drones and/or balloons. The overall system is divided into three main parts: preprocessing, vehicle detection, and tracking. In the preprocessing step, the acquired image gets undistorted and geo-registered against a user-selected reference frame. The current solution for detection and tracking utilizes machine powered by NVidia GTX 1080 cards.
We designed a highly optimized multi-layer detector and trained it on a hand annotated training dataset with 10.2M positive and 10.2M negative samples per class. The detector is capable of learning from its errors and improving over time. Due to the outstanding performance of detection and tracking algorithms, DataFromSky can provide high-quality data and guarantee a hit rate of over 96%. See the raw output from DFS deep neural networks below:
Our new scientific article answers… Since the inception of DataFromSky we were actively collaborating with academics to analyse traffic and design safer transportations systems for a better future. Beside the development of the processes to extract and analyse trajectories of the vehicles from aerial videos, we also aimed to analyse the accuracy of our approach itself.
In collaboration with Faculty of Information Technology and Faculty of Civil Engineering at Brno University of Technology, we are working to analyse the accuracy of object position estimation and accuracy of extracted trajectories and their properties when estimated by a low flying UAVs. (see our news from a year ago).
In the previous year, we have developed a tool to assess the accuracy of object position estimation algorithm similar to the one used in DataFromSky. We used this tool to estimate the accuracy of object position estimation from aerial imagery captured by a general-purpose drone in various scenes and compared the results with spatial data collected with an industrial grade GPS sensor. A part of this research and its results have been recently published in special issue on Unmanned Aerial Vehicles in peer-reviewed scientific journal International Journal of Transportation Science and Technology and is already available for pre-press preview at the following link: http://dx.doi.org/10.1016/j.ijtst.2017.02.002
The article provides an insight into the nature of the accuracy of position estimation and properties of uncertainty propagation through the algorithm with respect to various aspects of the camera, scene and its setup. The additional contribution of article is to provide a guiding tool to properly choose and set the drone pose and camera to achieve the desired accuracy of the position estimation of objects in the traffic scene prior the capture of the scene itself.
Picture: “Spatial visualisation of the resulting position estimation error in metres caused by non-linear deformation, across the camera field of view. The 4 red crosses represent the images of the landmarks. The camera is situated at position (0,0,100)[m] looking directly down.”Read more
Our team of DataFromSky comes up with lots of news, opening online webinars for our customers, potential clients and partners is one of it! We would like to get more in touch with You; therefore, we decided to start sharing our knowledge through a creative way via online web seminars, where everyone can contribute to the debate.
Agenda of DFS WEBINARS:
– Introduction of the project
– DataFromSky Viewer – software tool for advanced interpretation of traffic data
– DataFromSky Traffic drone & Traffic Balloon – special hardware for precise data collecting
– Practical demonstrations of application of DataFromSky in traffic analysis
– Research & Development – detection of near accidents, real-time traffic data
The webinar is opened and higly recommended for everyone, who is interested in learning more about DataFromSky! The only thing you have to do to join the online seminar, is to send your registration via email to the following email address: firstname.lastname@example.org
Please, mention the days, that would be suitable for You. The confirmation with the specification of the date and time block of the webinar will be sent back to your email afterwards.
One online webinar would take around half an hour and one time block is limited for maximum of 7 participants.
First webinars will start from 1. 3. 2017!
Do not hesitate and join our online web seminar to find out more interesting characteristics about DataFromSky.
We are looking forward to connecting on DFS WEBINAR!Read more
Flow analysis, speed and critical points in the city of Alba
“DataFromSky”, in other words, data that arrives from the sky through the shootings of a camera placed on a balloon. It was not so difficult to see one of these during these days in the industrial area located in Corso Asti, where the local study “TrafficLab” is going to analyze the traffic from the sky. “This is an analysis that we need for planning the new commerce building placed in this area” says the municipal officer. The results are still processing. Andrea Marella, engineer at TrafficLab, says: “We realized some aerial video at a height about 50/60 meters. The results are really interesting, because it is possible to see the traffic from a different perspective.” And explains: “This is a new methodology developed about three years ago that allows to analyze the vehicle movements from the sky. This service doesn’t interference with the sensitive data, and helps to get information automatically with our software. This can give us the number of the vehicles on the road, the traffic flow, the directions, the speeds and the conflict points.”
Looking for solutions
The camera is attached to a balloon inflated with helium. Marella adds: “We cannot use drones, because it is too difficult to have authorizations. We work in this way from a long time, with motorway companies, shopping centers and traffic local authorities and the results are very satisfactory. In this way we can have a wider vision that allows to report potentially critical situations and to find out how to solve them.”
DataFromSky in cooperation with COWI is coming with a new traffic analysis from Denmark! As in all other major cities, massive amounts of traffic goes in and out of Copenhagen every day. This leads to congestion and the everlasting discussion on how to effectively move as many people as possible. In Copenhagen, almost 50 percent of all commuting trips are made using bicycles. This results in numerous potential conflicts between bicycles and cars, and the battle for allocating and distributing sheer road surface space between the transport modes is continually on – especially at certain locations, where available space is scarce. One such place is “Christmas Moellers Plads”, an intersection complex connecting inner Copenhagen to the island Amager. Through the use of a specialized UAV and Data From Sky, COWI has obtained a complete 4-hour overview of traffic passing through this complex for both cyclists and motorists – and the interaction between the two. The data extracted now forms the basis for looking into new and better solutions that will keep this people mover perform at its best.