Add to Watchlist

GDAL - Process and Transform Raster Data


Citation of segment
Embed Code
Purchasing a DVD Cite video
unfortunately G. dollars a lot larger than get that doesn't go quite assembly of that and I'm also going to focus a bit more on Python bindings for judo and able to to see a little bit of a comparison in terms of complexity of some simple commands and then I'll have a bunch of links to different documentation so and to start I'll talk a little bit about what you know is and some of the basic functionality but it offers to some of the basic commands and a little bit about how to install it either for the command line or Python and a little bit about these accused and then some other resources and tutorials that already exist that I can recommend show all and is a very very brief history GL all have existed from whom since about 2000 but it was starting to be developed already in 1998 in about 2006 the open-source you spatial Foundation took it over as 1 of their own projects and has been developing it ever since and it was previously in were separated from what is known as the O. G. a simple features library which I'll talk about later but they've been much more tightly coupled over the last few years so you can talk about G all and when you speak about all often speaking about both OGI and Geitel and by most it is going to be focusing on the roster of functionality and the coal afterward will talk a bit more about the vector functionality currently were version 2 comma decimal 3 comma decimal 2 on as of September 2018 just to give an idea if you're looking at different versions to install
energy all itself has the promise of free software license MIT license the EC fairly frequently with different types of open source software and it's written in C and C plus plus that their bindings for python ruby Java Pearl C# and a couple other languages as well and it's cracked cross-platform which is really useful and not just Mac clinics Linux Windows but also Android and IOS the intended purpose of Jeong is basically just to read and write the RASTA and or vector geospatial formats and there's more than 200 supported geospatial data formats and and essentially presents the single abstract data model for both roster vector respectively for 2 different data models and that also provides different command-line utilities which will cover briefly like I said and mostly for translation and a little bit of geospatial processing but a lot of the processing is enhanced when you incorporate other languages with other libraries terms were process and modify the data in different ways it's most heavily not a geographic information system and so the RASTA dataset and you don't look like you basically have a dataset in whatever format and we created G doll dataset object and within that you have different objects that are each of the roster bands in each of those in different properties and that the dad itself dataset itself also has properties such as the size of the roster that you're referencing transformation projection information system different types of additional metadata that you can access and modify and like I said also for each of the bonds and so that's the basis for all the analysis and you'll see this and mostly come up in the Python bindings rather than the command-line utilities just some of the RASTA data
formats that are supported and there are many but you package is really interesting due to of course and then there's also this cheat all virtual RASTA that you can use that spacing XML file that links to the files that you have located somewhere else and you can use the virtual rosters to modify and conduct analysis without actually modifying the original files that's pretty cool I will be talking about that but something to look into that I would recommend the term also geospatial pdfs which are a headache and I hate them but you can tell without with them pretty easily all and if you ever run in 1 Fefferman 1 command lines commands to get the list of all of the formats has just too don't translate formats and you don't translate is 1 and said you know utilities I'll talk about more than that to some of the software that utilizes along with a lot of open source software that does that U. Jayess currently be due cattle or few toolbox sort edges and grass is mysterious and there are tons of software they used you how to deal with all the geospatial data that they handle and work with so what is the functionality actually
look like opened and you don't so here are just some of the utilities there's like 3 to 4 times more luminous and its little bit small but you know involved just reports the metadata and the dataset they are accessing jail translate is mostly used to convert between different formats but also to create copies GL work deals mostly with the projections and coordinate systems so and you can do that change those that also shifts data and new modified accordance and in different ways and there's another utility dealing specifically with the arms and you don't words you can use to mosaic different images which I use a lot of different and larger processing lots of satellite imagery for example gr roster iOS's and embolic ions link sort of the 2 different other models to translate between vectors and and it also can shop using using Geitel and and 1 also find very interesting is you can compare 2 2 images and report on the differences between which has output that you need to look at the documentation for more closely just an example this is
just the tool include yes or no and not so important but it's just the root projecting a dataset and we'll see down here and in a lot of the different tools a Q S is what is basically that she'd all command-line utility command and in this case it's too don't work because you read projecting agendas of learning more about some of the command line utilities and what the input output look at and don't wanna deal with all the documentation at 1st just for simple basics a lot of these tools will give you just sort of a bit of an insight with the user interface where you can play with some of the variables and look at what happens and it what he she has love the it's just it's just a regular Q gestural so this is just the regular re Projekt transform tools and you look at most of them will usually have out the command right there are a lot of people skip over it's like it's just a bunch of text and all had OK start the process but it's usually for a lot of the standard tools of science included in the window and until the connection that was there is you don't work was and 1 of the commands is then to overwrite the file that you already have with a new coordinate system in this case WGS 84 and to me then really list all the no data values of 0 the output file format is due to if you have an input data set in this case it was J peg and what everyone in Neenah output dataset it is said to be the careful with the endings for how you output down and if it doesn't quite fit with the output driver it won't necessarily not function but it might not open to this view that installation GL itself
is solid you have Q. J. us it's located somewhere and you want to explicitly in solid for you can do that using homebrew were around king chaos has a couple different binary distributions that you can access and install and Linux it's usually available through lot of the standard repositories that are available and the windows and there are a couple different class distributions that exist will definitely would recommend a list you know for Windows and it simplifies but had some of the difficulties that Windows 10's to try and for Python bindings citizen to
be talking a bit more about the Python bindings I'd definitely recommend using condoms regardless of what operating system you use and what kind of as I won't go into it now but it basically is a virtual environment package and dependency manager for Python and here's how to install it you can come back and look at the slides later if you're interested with a bit of a tutorial so
basic usage at all so the I did this image of the all and and had has a command of just allows me to show the different output images within the Dupre itself and the practice of using a very simple there's 2 Standard packed Python packages for operating systems and stuff which I almost mostly need to deal with past and now I for doing the radius which is the advantage to working in Python and just using Geitel and not limit for showing the plots and just you know for accessing formats and transforming things from the 1st thing you might wanna do is to check the versions of the things that you're working on this is general and you can check whatever version of Python that you're using with this command and whatever version of jade all that the version of iPhone you're doing is accessing with this command in this case it's 2 point 2 point to it also find out where it's located here and this just allows G also throw exceptions and I can purchase
not so important you can do is fix that the same information from the command line by just if you're Python installed it in the version this way and all season general info and this returns the version of G all that you have installed if you
wanna get information about a dataset and we'll use what's called the GED all info utility which I already talked about in the example that we showing that I just use a Landsat 8 bands to surrender and then also show a little bit of something with Sentinel to information this is the alpha that
you get from marginal info command and but the phosphoric yet so it
mostly just gives the driver that it is the file time where it's located at the corn information and exactly where the upper left to
lower left the corners of the image are located in accordance the projection it's in but also Latin long then this is done just with the G all info command in Python you can set the options for that command using a different object the command line a a lot
easier it's just you know and fell in the past the file in this case I also forced the recalculation of statistics to the maximum minimum and the different bands but that does the same thing as the other code pipe if you want an open and close a dataset in Python that's not something you need to do from the command line but basically discovered this is a link to the the lines of a panel when set band and and a closing just basically can deleted or set to so there is a function for that you can also access the metadata from the dataset that opened with a couple different commands this returns the number arrestors they're within the data set object and admitted some other information on what the new data values in this case there's none and what's the band datatype unsigned integer 16 and
overall statistics the maximum minimum of that and and then you can get a different way as well get maximum get and to create copies this looks at the difference in Python persisted all down 1st and showing a directory random directory on the path that I defined for its energy at the moment this to show that and that
opened the bands that I have I set the driver to whatever format I want in this case it's a virtual RASTA man and energy Dana and then in by the news that driver to create a copy and this is just where it's located and after that it happens very quickly you have the datasets than in the same directory and this looks also very different from the
command line and you don't need to deal with the opening copying he basically does is translate the kind of file said that you want and the input output data types it's very short and easy and how to go a little bit to this because
it's not so important and in the
code is available upon the slightest possible link that later so you can look at this a bit later and but if you wanna copy
image for example can you open it which creates said you know that as an object and you can define what those corners look like in different ways for the area with and that you want and I thought used translate to crop an image so use for copying we also use a 4 carbon and you can define the project the projection we know they're going to court to land you could do it that way to and from the command line it's they for
the same thing jail translated to the back
translated here and the projection window
which even has the same name in the same order the core that's the input data and the output data show it's pretty simple and I'll skip over
this but the longer example that goes into a bit more specifically of what makes the python
implementation of that more complicated than the command line to do different things and in this case I created a false
color image of using cheat on pipeline together and you'll see in all this
code here and you're accessing you'd all
data datasets and the data balanced in different ways and innovative ways that are and you realize that you'd all then model very explicitly
FIL is that agendas and from the
tutorials of Energy dollar-related has a longer list of a bunch of different things this cheat sheet in particular is very useful for command line commands I also highly recommend and this 1 the for different short introductions at looking at G. all in different ways from map projections to copying to just converting formats then there are other Python image
libraries so rest area I really recommend to action one work the fight on news and you know it simplifies a lot of the dealing with the dude all data models of datasets and roster and Blair's for this and and makes it much much simpler so you don't have to opening closing delete make sure that everything is happening and explicitly takes care of most of that for you and there's a couple different image libraries that recommend an and her some additional sources that I used was the documentation and at that prevents the attention that at the


Formal Metadata

Title GDAL - Process and Transform Raster Data
Title of Series Maptime Salzburg
Author Augustin, Hannah
License CC Attribution 3.0 Germany:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
DOI 10.5446/39178
Publisher Maptime Salzburg
Release Date 2018
Language English

Content Metadata

Subject Area Earth Sciences, Information technology
Keywords Geospatial Data Abstraction Library
Geospatial Data

Related Material



  392 ms - page object


AV-Portal 3.7.0 (943df4b4639bec127ddc6b93adb0c7d8d995f77c)