How long does it take to map an area with Mapillary? Well, apparently it’s up to you. According to he video below, it shouldn’t take long. It shows how street level photo coverage evolved in Phoenix, AZ between June 2015 and April 2016. I’m always amazed to see what can be achieved in just less than a year.
A year ago, high quality aerial imagery with a 10cm ground resolution was made available to the OSM community in Szeged, Hungary. It’s a very good example of not just sitting on the data but trying to make use of it. In theory, OpenStreetMap community can absolutely benefit from having a data source like this as there are way more details to be derived from such high resolution imagery. Also, the positional accuracy of the orthophoto is worth mentioning. You know, this is the kind of aerial photograph that you can make measurements on, like if it was a true map. It’s important because you can skip playing with different offsets and dragging your base map around to make it appear in its “true” position before you can actually start mapping. So, truth’s been told. It’s cool, but what the heck is with it?
Well, It’s been a year or so. I can talk about the benefits for days but it doesn’t really matter if no one is acting accordingly, right? There are things that “should” work in theory but when it comes to online communities… well, that’ a whole different story. Anyway, let’s lurk around and see what awesome mappers of OSM think about all this (oh, did I just say awesome people of OSM? Is it a spoiler? Oh well, I guess you have to click on the link below and read more to figure it out.)
Radius of gyration is a metric to quantify distributions around a center location. Its applications range from structural engineering to molecular physics. Since it incorporates the idea of dealing with locations, it can be applied for geographic data, as well. I recently came across with it in some global mobility studies where the goal was to characterize the travel patterns of individuals. In those papers, the metric indicates whether a person is more likely to travel long distances or not. In my research, where I am interested in geographic data contributions of volunteer mappers, I found it to be extremely useful to decide if the overall contribution shows local or global patterns. For many years, local knowledge was considered to be the main advantage of this so-called user generated geographic information. Local guys know the place, let them draw maps, let them take photos and the product will be accurate. While this is most probably true, it also seems that some of these guys like to do the same thing in distant places so there might be other factors than localness that can make these data sources accurate, therefore extremely valuable. Everything is up to the people who contribute, so the ultimate goal is still to understand their behavior. Now, enough of the crazy talk. Click on “read more” to do some fancy math and coding.
In my spare time, I teamed up with my old colleagues, the lovely faculty members at the University of Szeged, Department of Physical Geography and Geoinformatics to ask them for data donation. I couldn’t be happier to announce that after months of procrastination, some other delays and some more months of procrastination on my side, finally I am able to release a high resolution aerial imagery to the OpenStreetMap community for the solely purpose of mapping.
I wanted to get rid of long line segments in my data. Since I don’t know any software tools off the top of my head that would do the job, I decided to code it myself. I have already stored everything in a spatially-enabled PostgreSQL table, and to be honest, recently I am more interested in manipulating data at the database level to save some time. So, instead of writing a python script and looping through a cursor, I created a function which I am calling in SQL queries. The picture below gives an overview of what the following function does. Basically, it breaks input geometries at line segments that exceed a certain threshold in length. Red means “no bueno”, green means “yaay!”.
Today, a new project appeared on Kickstarter from the founder of OpenStreetMap, Steve Coast. I highly encourage everyone (in case you can) to stand for it so we can enjoy some insights about OSM. It’s pretty interesting how more than 1/5 of the goal have been pledged in less then 1 day.
In addition, enjoy a recent talk about the project.
As the last part of the previous post-series about MongoDB and Twitter I’m about to show some plots about an initial speed comparison of the two DBs. As a result, these plots show how MongoDB can perform better than traditional SQL solutions if it comes to speed. Of course the overall picture is more sophisticated. In these cases I focused on the simplest approach possible – retrieving documents from Mongo and rows from Postgre.
Originally I wanted to write about visualization in the 2nd post (and after that in the 3rd) but that post would have been too long to read. I always loose myself if it comes to writing but nevermind, finally it’s here. So, we know how to access to the DB and we can query for interesting subsets – even in a geographic way. All we have to do is to interpret our results. I’m presenting two ways, a gif animation and a wordcloud. It’s not about reinventing the wheel but still, I believe that these are useful approaches to complement each other.