Visualising Landsat data in QGIS

welcome to this video I'll be focusing on visualizing Landsat data in qgs so if you remember from a previous video where we left off of USGS where he had found this image here from the 25th of August 2019 this was a cloud free image or was killed and compagnon school and get downloaded and we had chosen to download all of these files here so yeah this I've done this and I have received a zip file that I have unzipped into a folder and I'm now ready to know this in huge ease so if I want to go to my browser and yeah you can see I have this full layer this was the folder that I downloaded and you can see that there in this folder is some text files that we can read some information about the files then there's a lot of TIFF files that are you can be perhaps remember from Oh Lancel image where we had the different bands let's see if I can switch over to finding them so here we had these different bands so we had band 4 which is the red band free which is a green band - it is a blue band so all of these bands are now represented as these small M TIFF files in my folder here so and four should be equal my red band so if I can normally drag it onto my canvas you see this somewhat hazy gray image does not read it's it's what it is it is reflecting how much red light there was reflected from the surface of the earth so where we have high values here zoom in here yeah here we have lots of red light this case you also probably have lots of green light and blue light so because these are gravel pits there'll be a lot of them of quartz that reflects may fish and in basic all wavelengths so this is our data it's also very hazy because it's not been what we call conscious stretching that's one of those things you should always be aware of when you work with image data is that you typically want to do some form of enhancement and the first thing you do is that you to stretch your image so if you go to properties of your lair and maybe we have this is shown as a single band gray and it has a minimum maximum but down here there's a minimum maximum value settings and the moment is set to the minimum maximum of the whole image what bibley want to do is that we want to say okay I don't want all of it symphony in image data that is the top one percent at the bottom one percent is noise and not very useful so we'll cut that off so we will focus on then 19 percent in the middle you can also do some mean value plus standard deviations I don't want to have it calculated for all of the image because basically on satellite data there's often some clouds hanging around in a corner or some data errors or something like it so we will have it set to this update canvas that means that each time I move the cameras around it will update the contour stretching so it will always try and stretch to what I see on my screen once that said note what happens to my image in the background it becomes much clearer because now it is stretching the image to those values that are in this range that I'm seeing here and if I now move around I don't want to do that let's take the hand and you can see if I move out into these that s where it's darker you see all the lights become much more lies enough somewhere where there's no land like here what happens is that it contrasts Thrax is that black area I see all of these patterns with spazzy is LG's in the water we can see ships and how the wake of the ship has stirred the water so you'll get some clean water from deeper up to the surface therefore we have these black lines after the ships so contrast acting means that in it ever you see now I'm over land it's completely white because I haven't let my mouse go yet soon as I let Mike Moscow it will contrast which to this area here so contrast stretching is really a rare one of those basic things you always have to do when working wolf image data oh we wanted a color image and if we go for so these are all our different bands if what we wanted to do that we wanted to have red green and a blue band so this case we have red for bean free blue too and we wanted to load them in into AM common image color image it's not quite as we would expect so you might say okay let's a dragon band free hand band too and of course what happens is that this puts the layers on top of each other and that's not very fun so what we have to do is that we have to do a little trick so I'll just get rid of my bands here and this trick is called a virtual raster what we can do is that we can avert Sylvester is a raster file that is reference into some existing files this the pretending that they are one file so what we want to do is really want to create one raster file that is referencing to all those raster images of us the layers and is the satellite imaging so I know I can do this type of thing up from my raster menu you'll be on dimension lemonis and build with virtual raster I arms always prefer to use the processing toolbox because here I can browse around and I can also find if I know that it's called operable virtual I can start typing some of the words and it will find that tool for me so in the virtual mapping Tulia we just a closer look of what that dialog box is saying so we have an input band we can put our layers in so here I can handle fula that's where I have all my image data and there's lots of layers here some of them I don't want some of them have come these aux files here they have come because I've had those layers are open already you see it's these layers that I loaded before so first of all I want to get rid of the to take file you and I want to get rid of all these Oxford's that I created before so get rid of those and there's a really annoying thing about this tool and is that once I load it but they will be I'll lose a regional name so they would this be called been one two three in that order they appear here so this one and one will be bad one but by band two will be bent in my band free will be band 11 so I have to make sure that my data is organized it's also noting that I have to be a k-12 that is if you remember from our slide here that band eight was a pen chromatic band of a high resolution the narrow bands and a virtual raster always has to have the same a solution it will take the average and I want to do too much manipulation so I only want to get rid of band eight or my collection so quite honest I'll just get rid of I hope we can move them down on this move and ten and exact I can drag the layers around so I can make the ordering so bad one and two and three and four and five and six and seven skip and eight and nine this is some quality assessment man don't want that there I've had the one that banned tone and nine and ten and eleven okay so I've got all my bands and listed up now my band these bands will get wrong numbers so I've just to make life easier I don't need them now I'll just also take them off so I'll just get my bands one two seven in their right order or the right names so okay to this hopefully and that's basically what I have to do oops we didn't do this right so if you got something wrong simply because you have a text file see I had a text file sticking around man one two three four five six seven yeah good so now it's fine and I've got my dataset and I'll disclose it and again you say this somewhat receive washy gray call everything so present what we want to change what you sees I'll go to layers and I go to its properties and here in I'm in symbology design what is my red band so if we remember so bread was banned for it is also named no and for here mine in a set so it'd be bad for bad free and bad too so red green and blue I'll make sure that my contrast stretching is again set to got 1% of the top there and I want to do it with my updatable each other and I'll say okay so now it changed to see what's blue now if I still very hazy because I'm displaying the whole image and there's probably some small clouds or a two hours hanging around that moved by my stretching but as soon as I swim in on an air like this you can see that the color stretching takes place and I am can now have look at by my images see you've eyes to it look in here it will stretch further until these airs down here and again as before if I move around so let's say I go out and look into the photo who booked here you know I can find I don't see that you can see again all of these different patterns from the algae in the water and you also see how the sand they fix here and we have the windmills and the fairies as before so now I have a what is generally known as a true color composite so I have combined the freebandz red green and blue to the red green and blue colors on my monitor is I ever not the only one often we look at our composites so you might look at where we have this near-infrared as our red and then the red is green and the Green is blue so this is typically these false colored ones are made to emphasize something specifically and this one is very good and identifying meditation and we will talk later about by looking at red and near-infrared is important when looking at meditation so our near infrared band is if we look at that is our so near-infrared is out here so that in our case is band 4 so here pop by and 5 so I was looking back in the Landsat 7 so then said 8 we have near-infrared in band 5 so if you've been sitting looking at been 5 4 & 3 so band 5 4 and free as air band combination so if we go into qgs and change the bands here so now look at band 5 & 4 & band free this will give me a as I said for scholar so the not doesn't look like if we were looking at looking down from the satellite or a plane it has some different but you can see that all these red colors emphasizing meditation I sleepy go and look a bit closer in cope hang here you can see how much more realization we have out and fights bear compared to a vegetable region so there's lots of tricks in order to emphasize special aspects that you'll be I just know I want to do is zoom out here that you might be looking for by combining different bands different ways to see meditation or urban spall or eat as we will be looking at in a moment so one of the key things to remember when we are looking at at satellite data is that it is data it's not it's not images it's not pictures of your wish it's recorded data and we can manipulate them one way of manipulating is to display them as a true color where we assign the red band from the satellite to the red color on our monitor the green from the satellites of green blue to a blue or we can choose as in this last case where I use the near-infrared and the red and the green as a bands I was looking at so satellite images are really satellite data that we can visualize as image nothing to remember when working with images and gugus that is in depend on where they are from satellites they're from airplanes or the drones or even being taken by hand camera is that you really often need to do contrast stretching so the thing about going into the properties and setting them in maps to some form of cutting off that the extremes to focus on very information is a typical important part of visualizing image data in qgs so this was this for this short video on how to visualize images in this case satellite Landsat 8 satellites and curious and later beam is we will start doing calculations and all our fancy finger so thank you for watching

Channel: Geoinformatics

Views: 3,665

Keywords:

Id: JA8v9a5eT7g

Length: 17min 42sec (1062 seconds)

Published: Tue Apr 14 2020