- [Steven] So there's no magic with this. There's no EXIF data. There is no reverse image search. - It's worrying that you
got so close so quick. Let's do something different. Let's see if you can
find an image of Nessie. (upbeat music) Hi everyone, it's David
Bombal back with my friends from OSINT Curious. Lizette, welcome. Steven, welcome. - Thank you, David. - Thanks David, good to be here. - So what have we got planned today? 'Cause I'm really looking forward to this. - We're gonna continue looking
at open source intelligence, but today we're going to
look at one particular skill and that is the art of geo location where you are given a image
from a photo or a video. And we have to figure out whereabouts in the world it was taken. - So this is really scary. I have given Steven and
Lizette some of my photos and they're gonna try and track me down. So I can't hide, I think
this is a warning, isn't it? For anyone who posts
stuff on social media. - Mm, maybe. (laughs)
(David laughs) Well, you posted a picture
of you hiding in a cave and then Micah, who was
there the last time, geolocated you quite quickly, which was a good hunch-
- A bit worrying. - For us to start again. (laughs) - So I mean, in this video,
you're gonna kind of... Are you gonna teach us how you do this? Is that right? - Yes. - There are lots of different
ways to geolocate an image and we'll look at different techniques, some are easier, some are harder, but there's a lot of
different ways to approach it. But the end goal is always
to find out where someone was or where an image was taken. Then try and verify as best
we can that it's accurate. - Okay, let's start with the first one. (upbeat music) - So here we have a picture of you, David standing in front of this
Egyptian kind of wall and looking at the
lighting and also the part that is visible on the right side, I could indicate this wasn't
actually in like outside Egypt. It was in a building
because I can see the lights in the ceiling. I can see that there are
other things there as well. So I can assume this might be a museum or at least an indoor area where there was a piece
of this Egyptian wall. So one of the ways to see
where a picture was taken is by doing a reverse image search. So a reverse image search
can be practically done by Google, Bing, Yandex and TinEye is probably the top four
where you want to go if you want to do a reverse image search. So a reverse image search
is asking the search engine if they can recognize an image, if it might be anywhere in the
data they have stored so far. So we're not searching with
words or letters or numbers, we're actually searching with a picture and it's going to be looking
at some of the fragments of a photo or a picture
to see if it can relate to something they have
already in their data set. And I like to use Bing
for a number of reasons. So when you're at bing.com
in the search bar, there are three icons
indicating on the right. And the middle icon is
the one I want to use because you have the option
to either put in a live URL, to upload a photo or to even
take a photo with your webcam, for example, but we're going
to be uploading the image we just saw of David. And now it's going to be seeing if it maybe recognizes the
photo from anywhere else. And the result we're
getting is a lot of pictures of people standing in front
of a beige colored wall with probably some encryptions in there like the one David is
standing in front of. But none of them are exactly the same as the one David is in, so
there are two ways to go. The first one would be
to do a visual search. So underneath David's picture on the left, we have a option called visual search. There will be a frame placed
over the photo and we can adjust the shape of the
frame and let it focus on other aspects of the photo. So we can maybe focus on one
of the pieces of the wall to see if it maybe will find
a matching piece of wall, but we're still not
getting the right results. Like all of them look
a little bit the same, but not exactly the same. And one of the ways to improve the reverse image image search is by cleaning up a picture and trust me, you don't have to do any
Photoshopping skills or whatsoever. Because there are actually
two great websites that can help you out. There is a website called remove.bg, which can help you remove
the background of a photo. In this case, then David
would be the person lining up, but we already know it's David. So David is actually the
person we want to remove from the photo. And then we're going to
be using cleanup.pictures, which is a online free tool. You can upload a photo to it. And as you can see on the example where you can see the paper
and the ruler and stuff, you can actually erase
stuff off the photo. Like if you think there
would be a great way, like if you have a photo of
somebody hiding behind a van, you can remove the van and
see the person behind it, that's not gonna happen. It's going to be using
artificial intelligence to color in whatever you removing. So we're going to be
uploading David's photo here. And now we're going to be
using the eraser function to erase David from the photo. Try to be as precise as
you can be, be blurring. There we go. It makes a guess of the color
that would be behind you. But now we have a full image of the wall you were standing in front of. So now if we take a screenshot
or we download the photo, as we see it here, there we go. And then we go back to Bing and we're going to be uploading it again. So now the first image we're getting, we already see that there are like three or four significant like
bricks and we see them in the first photo as well. And if you now open up the first result, and now we have the exact
same wall David was standing in front of, and now we can find
out where it was taken. In the purple, it says description. And at the end, it says in Oxford. If you go to the Flickr page,
it has the exact museum name. So that's how we found it out. So at first we did this
reverse image search where David was still in the picture and Bing was only showing
me other people standing in front of a beige colored wall, probably with Egyptian figurines on it. But it was not giving
me the exact same thing. And sometimes you need to clean
up a picture in order to get to that result where you were looking for. So in this case, we removed David without any Photoshopping skills, just by using a free tool
available on a website. And we were able to locate
where this picture was taken. And that's a very relatively
easy way to geolocate something by using a reverse image search. - I wanted to ask you,
can we look on Google? If you go Google Maps, can
we look in the Ashmolean? 'Cause they've probably
got it right there. - Yeah, let's have a look at Google Maps, places like museums and
places of public interest have a lot more additional photos in them. Okay, so here is the
Ashmolean Museum in Oxford. If I turn on Street View, you see, we've got all
these photos spheres on the inside that people have taken. So these are all fixed,
so we can't move around. But... - Yeah because some of
the museums allow you to walk around inside of them, right?
- Yeah. And this one doesn't seem to, let's see though, you can figure out. Now this looks less Egyptian,
this part of the museum. - Yeah, obviously here we're relying on someone taking a photo aren't we? - Yeah, if there's one in there. - Although sometimes the museum, they post what kind of
collections they have and they sometimes put-
- That's true, yeah. - Photos of the collections
they have as well. And we could maybe find
the image there as well, but I haven't... In my search for this,
I was just really happy to find the exact museum where it was. - Google, you have obviously
Google Street View, but you get user uploaded photos as well. So when you click on the location, so if I select the Ashmolean Museum, I can see these photos here
on Street View and 360. So these are all the 360 views. - One of you type the Ashmolean Egyptian, they might have the... 'cause the Egyptian thing is
normally there quite often. - We'll look at their website and then... Oh yeah, we've got loads of 'em. This looks kind of promising. So you learn so much in open
source intelligence, right? So what's this? What are the galleries
on the ground floor? Right, I'm trying to
see if it has something like, you know, a 3D view
or something like that. So it's actually got quite
a lot of Egyptian stuff. - I found it, I put it in
a chest, it's not a 360, but it's the same wall. - So this is from a review,
looks like a review website, Amazing Days Out. This is the same thing that's
seen from a different angle. I guess that kind of helps
with the verification aspect of it as well. So we had our hunch about where it was, but we actually need to
prove where it was (laughs) so we can use other sources to do that. - So if you go to the original photo, I thought behind my knees,
there's a sign there, I thought that would be a dead giveaway. Did that help you as well? Did that kind of like hint towards it? - No.
Or did you just realize, "Oh, straight away it's a museum." - I realized straight away it was a museum because we can see that
there were other areas like on the right where
you can see the lighting and it looks like museum-ish. Sometimes you see these exhibitions at airports, for example. But for me it felt more like a museum. Googling for where in the world museums have Egyptian expositions or where they have
pieces of wall available. That could be one of
the paths to go as well. But I thought it would be more fun to do a reverse image search-
- Much more fun. - And to figure out if I can find it that way.
- That was great. That was a great example. I really like that, cleaning up and reverse image thing was brilliant because that's not something
I've seen people do. - I've solved like 90% of
the photos you've signed up upon by doing a reverse image search. It was like really easy. So for this picture, for some reason, I assumed this wasn't
the UK, 'cause for me, the sea was like too blue and
the skies were too bright. - That's right.
(laughs) No offense to all of the UK people. But on the other hand, I've never been to all of
the coastlines in the UK. So there might be places where
the sea is as blue as it is on this picture but I had my doubts. I guess it would be a lighthouse or a viewing point
somewhere because it looked to be higher than the ground level. And of course there was
the coastal line as well. So one of the ways to do this is to like look at
every single lighthouse, somewhere along the coastline
to see if it would match, but that would take you
years and years and years. I took this image and
I went to Bing again. And the reason why I did so
is because I would be able to have the visual search. So to have the frame that
is overlaying the picture and I would be able to crop the picture to whatever it is I
want Bing to focus upon. Bing is showing us a lot of similar images of people or families. There are most of the time, two or three people in the
photo next to what seems to be a fence or a railing
overlooking a coastal line. But none of them are matching
with what we are seeing here. So I did a visual search and
I've focused on the red pole. - It's probably fair to
mention that when you crop an image like that in live time and it searches again for you, it's more of an art than a science. Like even like a few
millimeters either way can cause the computer
to find something else. So it's worth-
- Absolutely. - Trying a few different angles. I've just moved it there a tiny little bit and suddenly get all different results. - I think that was the one,
I think it was mirrored, but I'm trying to see my... Yeah, so there we go. - All we did was just cropped image just a few pixels more to the direction. And we found it.
- Yes. So I had this image similar
to this one as well. And when looking at the image, I saw that the top of the pole, the shape, like with the little cone and it was like a square cubicle box. I noticed that those were the same. So I knew that if this
picture didn't take me to the exact lighthouse, it might take me to a lighthouse that was
built by the same architect because the exact same railing was used. So at least it would give
me a hunch of the location. So in this case, if we
click on the visit site, so we now have Agulhas and Struisbaai. And this is the exact
same website, by the way. So we see here the railing and we see the same kind of shape. Well, I don't know if it was a advantage, but I've actually been to this location. I didn't recognize it by the way, 'cause I've never been up the lighthouse. But I know Cape Agulhas has a lighthouse. And it also explains here in the text, it says that, "This is
a small but famous town, home to the Agulhas
Lighthouse, monuments." Et cetera, et cetera. Then I Googled for the Agulhas Lighthouse to see if indeed the rest of the picture would match as well. And while I placed myself
there with Google Street View and looking at the photo
spheres that were taken there because Google Street View
doesn't go up the lighthouse, but a lot of people went there and they actually took those
photos, which we saw earlier, which Steven was showing
you at the Lake Loch Ness. And you would be able to
either have a 360 view or the opportunity to
just view a still frame. And that is what actually
helped me confirm the location. So if we now zoom into
the little red drop, we see a couple of little
blue balls lining up and we have to just check a couple and one of them was taken right up on. There we go, we see the same railing now. - So these people were
like a level further up on the light house.
- Yes. - Yeah, you see them down there. - And then by looking at this image, we can see the coastal line
and we can see the houses that are in front of the coastal line and we can match those
up with David picture. - There we go, so that's
half on the rocks here. And there's the path in the
rocks, it's a little bit- - And we can see that
there's a little fence around the white house on the corner left. And that fence is also shown
on David David's photo. - Okay, so this helps with
the verification part, doesn't it? The bottom left of the
image here, it's there. So yeah, we could figure out
we are within a few meters or even a few feet of
where David was there. - Yeah, I reckon that photo was just... Or that Google thing is just
above where we were standing. - Yeah, as an estimation. And of course I've been doing this for quite a number of times,
so I have some experience. This took me about 17
minutes and it was pure luck because as you guys just saw, looking at Bing using the visual search, is really comes down to
just moving it a little bit and moving it a little bit back and you suddenly get different results. A part of it is pure luck,
a part of it is experience. - I wanted to say this, if you go to the neighboring
town's page and just go back to where we did the photo searching, so like the third one on the top, this is me obviously knowing exactly. It says Cape Agulhas there and
if you look at the railings, it's got the silver and
the red kind of like, if you go back to the original photo. When you were talking about the railings, I noticed the silver and the red, it's kind of like exactly... It looks like they just
took it a little bit further towards the coast. So we was probably standing
a meter away from there. - Absolutely, but what we're seeing now is the cut Stephen made in the photo, which you shared with
us is slightly different from the cut that I made. So on the screenshot I
took here is the result that I got. So when Steven was making the rectangle on the visual search in Bing, we got a lot of railings,
but on the one I tried, I only get one result where
the railing was completely red. So David indicated first at the results, Steven was getting, sorry, that there was indeed a railing
where we had silver bars and we had red poles. But as you can see, the moment I was actually
doing this search for the first time, I only get one result with only the red railings
and the red posts. And this was for me, a great
way to see if it might be the same lighthouse or maybe
that it was the same architect. So I just wanted to share with you that it can come very
narrow to how wide or small or where you drag the rectangular
icon in the visual search. - A tiny change in cropping
can make a big difference. It's probably worth mentioning that the main different search engines. So you have Google Images, which incorporates Google Lens now. Obviously Bing as we've
seen and Yandex Images. Sort of the big three, they all give different
results for different images and they all behave differently. So if you don't get a
result that you want, or that you're looking for
with the first search engine, then try the others because
they all give different results. Throw in cropping, and you can
get different results again. - If you want to speed up this process, so let's say you've seen
this video and you think, "Well, I want to do a
reverse image search." On specific images you find
online, there are add-ons, which are little software
pieces you can add on to your browser in order to make the OSINT life easier for you. I really like using RevEye, which is available in
Chrome and in Firefox. And just by right clicking any image, it will give you the opportunity to search for that particular image in
four of the search engines. So it searches in Google,
in Bing, Yandex and TinEye. So TinEye is a search engine only for reverse image purposes, the results aren't, in my experience, that good as I've seen with
Google, Bing and Yandex. But sometimes I get a result in TinEye, as I do not get any result
in Bing, Yandex or Google. - There's a tool called Search by Image, which is free and it runs
in Firefox and Chrome. So you get the add on for your browser from Firefox Store or from Chrome Store. But it just sits up here
like a little camera. And it works in a couple
of different ways. If I have an image here, for example, and I want to do a reverse image search, I don't have to download it to my computer and then re-upload it or paste the URL or anything like that. If I right click and then
search by images installed, it will present me with different options for different reverse
image search engines. I can just click on Yandex, for example, it will automatically upload
that image to Yandex for me. And I'll get lots of different results, which may or may not be
what I'm looking for. In terms of workflow, when you're doing that many, many times, it makes life a lot easier. But you can also crop with
Search by Image as well. Click on the period at
the top of your browser. There's a, what it
calls a capture feature, which is the cropping feature. So I click on capture then
choose which search engine I want to search with, so I
could choose all of them. But to stop clogging up my browser, I'll just choose, let's do Google. And then I can select the actual image that I want to search here. So if I just want to focus
on one particular detail, let's say this post design here, I'll put a couple of them
in and then click search. You'll take that little
cropped image I've just created and pass it to Google Images. And then Google will show
you what it think matches. For regular data, Google is
probably the best search engine or is the best search engine really. But for images it's not
as good as Yandex and Bing for looking at exact matches. But Search by Image just sits
up there in your browser. It just from a workflow point of view, means you can do something in
two or three clicks that... So you don't have to download
an image, re-upload it, copy paste it, things like that or crop it in Paint or Photoshop, it's a time saver, especially when you are
doing a lot of these. - Google is improving their
visual search with Google Lens. Although I must say that
I'm still not as satisfied with the results as I am getting them with Bing and Yandex, for example. So Google is stepping up their game, but for me there aren't
there yet. (laughs) - Google Image Search and Google Lens were two separate things. So Google Lens, you used
to work on your phone. So you could point your phone at something with the phone's camera and
it would tell you what it was. So you could point it at the
flower and it would tell you, "Oh, that's a daisy or that's a daffodil." It would tell you what it was. Only in the past few months, I want to say, maybe in
the last six months or so, you can access that feature
in the Chrome browser now. So again, you right click and it says here, search
image with Google Lens. That used to say search image with Google reverse image search. And let's see, it works
slightly differently. And again, we can do that like little cropping thing in
here like you could with Bing. - And funny to see is that as you see now that we did a reverse image search of the cropped parts
just in the Google Images and it didn't give us such good results. As we were using Google Lens now, the results are improving a lot. 'Cause we can actually
see that we can recognize the Cape Agulhas title again
in all of their pictures. - (laughs) Yeah, same
company, but different tool, even though they're very similar. And although there's no text here, you can extract text from images as well. And it will translate text in images, if you are working with foreign languages. (upbeat music) - Lizette, let's put you to the test. Here's a photo.
- Okay. - Let's see how long it
takes you to find it. - Okay, so first I would
guess it would be somewhere in the UK because you can see the chimneys and the whole building feels to me like it might be UK-ish, Scotland. I don't know somewhere on
your beautiful islands. I would take a reverse image search. I'll probably start at
Bing because I can see some of the areas I might want to crop. So I would go to Bing, upload the photo and let me see if there's
something similar. - Okay. - Yeah, I think we already have it. Yes, that one. - This one? - Can you click that one? - Yeah, this is gonna take
us to a video, I think. Oh, but it's given us a location there, Charlecote Deer Park
and there's a link here. - Can you Google for Charlecote Deer Park to see what kind of images
it gives you in Google? - Okay, yeah. Okay, so Charlecote Deer Park, was it? - Yes. - Charlecote Deer Park. - And then go to Google
Images to see what kind of... - I will even spell deer
correctly, how about that? - (laughs) I think your very first photo on Google was actually
the one that was right. - I think you're right, yeah so... - What does the stopwatch say? How much time did I spend? - That didn't take you very long. And we were wasting time in between, but I mean, that was under a minute. - So something so easy, like a reverse image search can take you up to like three minutes
to guess where somebody is. Even though on the picture, none of the things except
for the building style, could give away where it was. We didn't see any street signs. We didn't see any license plates to guess which kind of country it was. Anything that might give us a direction in what kind of country it was. So just a simple reverse
image search will give you the result where somebody could be. (upbeat music) - So if you look at
this, it looks difficult. I would've thought of,
you know, just because... I mean, there's not much to
go on except the number, yeah? - Yeah, so I looked at
the image and I was like, "This is probably a bridge." Because in the upper right corner, I could see that there was
like a top side of the bridge, 'cause there was like air
behind it and we could go under. It wasn't like a big river, so I assumed it was like a
small canal maybe somewhere. I knew this wasn't in the Netherlands because it doesn't look like the style of build we have here, where I live. So I just simply Googled for the words, canal 200 bridge. And then I look at the Google Images and it's the third image. - (laughs) Just like that, yeah. - So even though there didn't seem that there were a lot of pivot points. Like if the number 200
wouldn't be above the bridge, my first guess would be to
do a reverse image search to see if I can find similar images where I can find a similar
style build of the bridge. But in this case I was like, "This is a bridge, there's the number 200, it's a small canal, I'm
just going to be typing in three words and see where it takes me." And I was there in less than a minute. - Okay, I'm gonna try
to reverse image search, see what comes up. Even though this isn't the exact same one, it would tell you... Like you even figure out
roughly which canal it was or whereabouts in the country it was. So there's that similar bridge
that looks like number 80, that's number 41. So you probably would
figure out where that was. Let's try Yandex reverse image search. So Yandex is kind of a... It's usually my first port of call. - I haven't been since the war. - Oh yeah, of course, yeah. - Is Yandex in the Ukraine or is it in Russia-
- No, it's a Russian-based website. - Okay. - Yeah, so they are
Russian based and owned and they certainly with
their regular search results, they filter them to be... So results tend to be more pro Russian although that's presumably
less of an issue for images. But yeah there's no like
exact spot matches there. - Do you know what would be interesting, could you show it in Google? (keys clacking) So just like Bing and Yandex, Google has the option to do a
reverse image search as well. Where we can upload an image and Google will do the
reverse image search. Other to what other websites do, Google actually looks at the image and tries to identify
what it sees on the image. We can see that after the image title, we can see that Google
indicated this picture as being leisure. But you can remove that text
and make Google focus more on what is happening in the image. So we could example type in, bridge 200 and see if the results would be improving. I think the first visually
searched is a similar image as well as the first blue URL are the ones we are looking for. - So it's bridge 200, Station
Road Bridge number 200, Hayes Bridge 200, yeah. - That's right, yeah. - Heyford Common Bridge. - Well... - Shows how differently
each search engine behaves in terms of what it
thinks you want to see. And they're all very, very different. - Let's do something different. Let's see if you can
find an image of Nessie. And it gives it away though. It was a bad photo for
me to give you guys. - It was yeah. (laughs) So this is the first photo you gave us and we're gonna see
how we can find exactly where this was taken. Now I think we should
mention first, in fairness, is that one of the most
common ways people think about geo locating an
image is with EXIF data, which is where the GPS coordinates of a image are embedded in the photo. And there are a few
different tools you can use to extract that. But the reality of doing
internet investigation is that you won't find EXIF data on most of the photos you find in the web. Because all the big social media platforms and so on will strip that from an image. So it's good to know it's
there and it's really useful. It's very accurate, as long
as it hasn't been faked, of course, but most of
the images you work with in real life will not
have that data there. And it'd be quite a
boring webcast, I think, if we just sat watching us
extracting EXIF data too, right? So... - And it could be a
red herring, the photo, I was just thinking, how do you know that I actually took this anywhere near where it's supposed to be taken? - Well, exactly. And EXIF data can be edited so you can change the GPS
coordinates in the image and make it appear to be
anywhere in the world really. But that's why it's important
when we find somewhere that we have to verify
that it's right as well. So there's two parts. This first part is finding where it is. And the second part is actually checking to see if it's right. So we don't just guess. - I took that photo in London. - (laughs) Yeah, the idea... I don't recognize that part
of Hyde Park, but maybe. How we want to approach
this image is to try and extract as much
information from it as we can. So we would look at the
background, the foreground. So we'd see the hills in the distance. There's this huge lake, which obviously dominates the picture. There's a valley off
to the right hand side and there's rocks in the shoreline. So what you cannot escape
in this picture, of course, is this monster, this dinosaur,
the Loch Ness monster. So there's a bit of a contextual clue and if Loch Ness monster
was not in this image, it would be harder but not impossible. It would just take more time to find out whereabouts this was taken. And as you'll see, Loch
Ness is a big place. So actually we are going to
find where David took this to within a few meters, just using some observational techniques. So there's no magic with this. There's no EXIF data. There is no reverse image search. And if you don't know what
a reverse image search is, we'll look at that with
some of the other photos. We wanna first of all, not just
assume that it's Loch Ness, but actually check that it is by finding Loch Ness on the mapping system and comparing some features
we find in the photos with what is there in reality,
to see if things match up. What we're going to do is find Loch Ness, which we is our hunch, where this is. And then look at the
features in the photo, and then see if we can
match them to something. So there's a few different
mapping resources we might want to use. Google Maps is obviously very well known. There are other mapping
tools like Bing Maps. But I'm going to introduce a tool that you may already be familiar with, that's called Google Earth
Pro, which is free to use. It runs on Windows and runs
on Mac and runs on Linux. Basically, Google have created
a 3D model of the world and overlaid it with years and years worth of satellite photos or aerial photos. But also with terrain information as well. So that allows us to
recreate a landscape in 3D, which will be really
handy when we're trying to compare our photo here. 'Cause if you look in the
background of this image, we have some really
distinctive shape hills, especially the lumpy one in the middle and that's quite distinctive. So that will hopefully help us. So this is Google Earth. There is a browser based
version of Google Earth, which doesn't have the
same level of features, but this is the desktop version. So we'll find Loch Ness and to do that, we can type in up here Loch Ness and then Google will tell us where that is and it will zoom in for us. So this is Loch Ness but you
can see it is quite a big lake, certainly by UK and European standards. But we want to be accurate
to within a couple of meters as to where David took this photo. If we were trying to find
where David was in real life, we want to be as accurate as possible. So down here, there are what
Google calls, different layers. These are different features
that you can turn on and off. So we can turn on and
off photos, for example. And if we enable photos, we can see images that people have taken
at locations around here. As you click on one, you get more of a ground level view there, which might come in handy. We can turn things like roads
on and off, if we wanted to. Google also has a 3D terrain feature. And when we zoom in, we can move around a little bit like this. We can actually see if I go low enough that it renders the terrain in 3D here. So I'm probably just a few 100
feet off the ground, I reckon but you can see, we have this 3D effect. So when we go back to our image, we want to try and be able
to look for these features from David's photograph and
match them as best we can to the landscape around Loch Ness. If you look at it from here, all these hills to someone who hasn't... Like me, I've not
actually been to the area. I don't know, Loch Ness particularly well, all these hills look kind of the same from a bird's eye view up here. So we want to try and find something that is as distinctive as possible. And the most distinctive feature
on here is this hill here. I guess you can see we have
all these dozens and dozens of hills, but we want one
that really stands out. The more distinctive something is, the higher chance we have
of being able to find it on a satellite photo database, like Google earth or
another mapping program. So what I want to be
able to do is to recreate in Google Earth, a photo in Loch Ness with this huge hill in the background. There are a lot of hills
around Loch Ness, for sure. - You said you've never been there, yeah? - I have never been to Loch Ness no. Scotland is a beautiful country,
I've been there many times. But I've never actually been to Loch Ness. - Kind of a cheat if you know the area. - Yeah, absolutely. And in geo location, like sport, there is such a thing as
home fixture advantage. Like if there's some
areas, you know well... Although as Lizette will show us actually with one of the other photos, just because you know an area
doesn't always make it easier. But in this case, yeah,
I've never been here. So because we can do
things in 3D, I can just, by holding the shift key and
holding on the mouse button, I can rotate myself here to
try and recreate the view. So I put myself right in the middle of Loch Ness as best I can. And if this isn't right,
we can move around until we find what we need. So zoom down, if I go too far, it'll like pitch me right
into ground view level. So this is pretty much ground view level. And then I can just hold the shift key and the mouse and spin round. So if I look this way, so I'm looking kind of north east there because there's a compass in the corner, which will tell me where. Looking that way, it's kind of flat towards that end of the Loch. And if we look at the distance here, it's actually gets hillier in the distance rather than flatter. So that tells us, we
are probably not looking in the right direction. We'll pivot round by 180 degrees or so. And look at the other end of the Loch. And you see this is a
little bit more like we see in the photo. When you're doing this
on your own, by the way, it's really useful to have two monitors because you can have
the image on one monitor and your tools on the other. It looks more likely that
we're looking this way sort of Southwest-ish along
Loch rather than Northeast. The thing I want to look for and you can kind of see
it sticking up already is this really distinctive hill. And from ground level round Loch Ness, there are no other hills really that match the big lumpy hill. So switch back there. You see how distinctive it is standing up above the lakeside. So we wanna say it's
probably this hill here because there are not any
others that really stand out. So we want to guess it's
probably this location, but guessing isn't really good enough. We have to try and verify
and be as certain as we can. So to do that, we will pick
out some other features. If I stand where David was
standing when he took this photo, looking straight ahead,
I can see the big hill, which we think we found. To my right, I should be able
to see this V-shaped valley on the other side of the lake. And if I'm right, I should
be able to see a valley to my right here, the
angle's not quite right. So I probably need to... I'll zoom out a bit, I probably need to move back a little bit. And if I come here and
then I'll zoom back down to ground level. Actually that looks a
little bit better there. You see we've got the hill there. So which is in the center
like in David's picture and we have this valley to the right and that is not bad. Let's have a look. So we got a valley to the
right in the correct position. We have the hill in the correct position. We have the Loch going off to the far left in the distance on this picture. So we should, if we're right be able to see Loch in the distance,
the forest on the hillside, the big hill in the middle
and the valley on the right. So that matches and you have
to say, except for one thing. - Very impressive.
- Yeah. (laughs) We know you weren't standing in the middle of the lake
when you took the picture, you're on the shoreline. So... - I was in London, I told you. - (laughs) What we want to
do is try and and say, right? Let's put ourselves in David's
shoes as literally as we can and figure out well, like,
unless you are on a boat or something and it doesn't
look like you were actually, 'cause this looks like
shoreline and there's waves and looking at the model of
Nessie, it's not very deep here. So we probably will assume that you are at the edge of the water. We'll probably say you
were about here somewhere on the shoreline, there's
the main road here. We can turn roads back on actually. In fact, if I enable
roads, see it was the B852, which is one of the roads
alongside Loch Ness. So you are probably, I guess
round about here somewhere. This is a photo that someone else took at a very similar location. You can see we've got the same hill, more or less the same view to
within a few 100 meters or so. So if you were a fugitive or something, or I was wanting to
find out where you were, I will probably say to
within a few 100 meters, you were along this bit of road here and I could copy out the
GPS coordinates from here and load it into another
tool for some more analysis. This, I would say without checking looks like maybe a car parking
area or something like that, where you would have to park up. But yeah, we'll be confident
of your position there just by looking at the
features in the photo and comparing them to ones
we could find in the area. No GPS, anything like that. And for your viewers, we have
not pre-consulted on this. You've given us no clues
whatsoever, am I right? (laughs) I hope I am. - I would say so, yes, definitely. It's a pity you can't look on Google. Can you put the... What's it called? The man on the road and see, I dunno if it shows anything. - We can do that in
Google Street View, yeah. I should mention that if I get low enough to the ground here, if there
is coverage here, it should. Oh, there we go, should
put me in the Street View. There we go, it's just a while to load. So this is now Google Street View, which is Google's collection
of images that's taken from cars and things like that. - So if you move towards the right, like pan round to the right. - Yeah. - I think there's a foot path right down the side there, yeah, there you go. So if you go a little bit forward. - Okay. - I dunno if you can get like
a top down without the 3D, you could probably see
exactly where we sat. - Cool, yeah, so... - Didn't take you long, wow. - Yeah I think... And people will look at that and say, "Okay, well the Loch
Ness monster is a clue and it gives you the lake." Although Loch Ness is a pretty big area. If Loch Ness wasn't there,
how else could I find this? Well, it would definitely take me longer, but you'd want to look
at the terrain and say, "Well, it still looks
like the British Isles, and possibly Northern
Europe, somewhere like that." But you would say, "Well, this isn't Asia. It doesn't look like north
America particularly." So you would come up
with a list of big lakes and then try and narrow it down, which would take a little bit longer. But the principle would
be the same for sure. - Can you go to Street
View just in your browser? 'Cause I think... Sorry to take you off track now. Let's see. - So this is the Loch Ness monster and I'll you show something I noticed when I was preparing
this, it's really cool. You see the little orange
peg man in the corner here, so you normally click on
to enable Street View. When you get close to Loch
Ness watch what happens. Like it turned into... So it's turned this like
a little green Nessie, their little Easter egg from Google. So we click on the Street
View icon and all these areas that are blue are areas
that have been captured by Google Street View car and
all these little circles here are user uploaded panoramic photos, which are really good for verification. So I think we were about here somewhere. In fact, let's put the satellite layer on, that makes it a little
bit easier to work with. Here we are in a place
called Loch Ness view. Here we are, so these are
the little photos spheres that users upload. So this is actually pretty
close to where you were, I guess. - Yeah, there's no Nessie
'cause we put it there. But I mean, (laughs) no, I
actually saw the real one. Come on. - You would not be the
first person to see it, but then be... 'Cause we struggled to verify it. So actually when you look
at where we are there, where we worked our way
to, I look at that there, like you got the mountain,
you got the valley. It's pretty good. - That's probably exactly the place because the rocks look
very similar, don't they? - With these one, with the photos spheres, you can pan through 360 degrees, but you can't move, unfortunately. Whereas the ones that were
taken with a car you can. And in fact, this one looks like someone's been on a boat, right? The Google Street View boat. Okay, so we can get... This is quite rare actually
for people who play geo guesser and things like that to have
the Google Street View boat, as it were, I guess that
we are looking there back towards the shoreline. So you can see-
- I've never seen the boat. - It's normally a car or sometimes a bike or even someone with a backpack, if it's a hard to reach area. But I've never seen the
Google Street View boat. There we go. - I can't remember exactly where it was. It's either that Loch Ness view or the one like right next
to it that we looked at because there's a few of them that are right next to each other. It's worrying that you
got so close so quick. - That's just from finding
features in the image. - Could you show the EXIF data
and we can just confirm it? - There are many tools which
will let you view the EXIF data of an image. This is one I'm going to
show just because it's here. And this is a tool called Forensically. It has a few little like
photo inspection tools, but it has its metadata one here. So I will upload the file
from my desktop, this one. And if I click on the metadata tab here, you can see it has all this
extra information in there. So it tells us the make
and model of your phone, which is an iPhone 11 Pro Max. The date and time of the photo. So it was the 28th of August, 2020, but we have all this GPS
dat embedded in the image. So we have latitude, longitude, the speed at which you're moving. If you're moving even slightly,
which way you are facing. And some of these features are fairly new. Like you wouldn't have had
these on phones and devices even a few years ago, GPS image direction. So it gives us a bearing that
you are actually looking at. And there's all sorts of other information about the photo too. So this is really, really
useful information. And if I click on the
geo tags feature here, it will actually pull the tags out and put them on the map for me. There we were by the side of Loch Ness. View on Google Maps, let's do that. This should be accurate to
within meters, centimeters, even. There you are right on the shore. So yeah, there's a lay-by
which I guess you parked in, let's go back to Street View, yeah. So you park in that lay-by there. So this photo sphere here is literally about two, three meters from
where you took your photo. - Yeah, I think that rock that's
just underneath the photo. If you pan right down is the rock that you see in the picture maybe. - This photographer was stood on... That is probably this rock here, I guess. So yeah, you can see the
big hill in the background. You can see the valley
is a little bit hazier, but yeah, that's from the GPS data. It's good to know that GPS data exists because it's a shortcut
to finding a location. But the reality for most people
that deal with this stuff is you are not going
to get images on there. If you upload that photo
to Twitter, for example, all the GPS data that we've just looked at will be stripped out. But just because you don't have EXIF data, doesn't mean you can't find
where things are, definitely. - How close were you? You were like within a
few meters or very close. There's so many of those little lay-bys. It was either that one or
the one right next to it was the one that you chose. - Yeah and if I wanted
to know where you'd been, for example, 'cause I was spying on you or you're a fugitive on the run and you posted photos on
Instagram or something. Like that would be enough
to get your location to the right area, definitely. With patience and with
tools like Street View, you can get a very high
degree of accuracy for sure. - The good thing about
social media, like you said, is the EXIF data is stripped out. Just for everyone who's interested, I put a link to another
video that I created, where you can actually go onto Flickr and you can find my photos and
you can find out where I was. You can do that as like a test, if you wanna do some EXIF data stuff and find out where I was on other photos. - Flickr is a good place
to practice actually, 'cause it's dedicated to photography. So it leaves all the EXIF data in there. So yeah, there's lots of good stuff to practice with on there. But yeah, Facebook,
Instagram, Twitter, LinkedIn, unfortunately, all that data is gone. (upbeat music) - I was going through your photos and of course I geolocated a bunch, but this one got my
attention straight away because probably everybody
who sees this probably says, "Well, this is Amsterdam." But how can I be exact? How can I know for sure it was there? So I took this image and I've been to Amsterdam many, many, many times. But I couldn't recognize
it straight away saying, "Well it's probably at
this canal specifically." So this image was quite fun for me to do because I had to focus on
a lot of other pivot points in order to confirm that
this was indeed in Amsterdam and exactly where it was. So first things I looked
at was the light pole on the bridge, it's very
typically for Amsterdam. So I Googled it just to be sure that these were the light poles that are used in Amsterdam as well. On the left, you see a parking meter. So I know in Amsterdam
there is paid parking and actually all of the
parking meters have a zone code in the white little
rectangle above the letter P. I can zoom in all I want, but I can never read
what numbers there are. There might be a one and
a two or a one and a three at the end, but it doesn't get too clear. That's a pity because in Amsterdam, if you Google for the parking zones, it will light up the areas where this parking zone is active. And I would have a very narrow
area where I would be able to go on, for example, Google Street View, to see if the location would match. This was unfortunately
not my best pivot points. So I focused on the houses. A lot of the canal houses
in Amsterdam are monuments. And you can recognize
those by a little label, which is often next to the door. But those labels can often
be mistaken by family names of people who live in those
houses or house numbers. I tried to see if I can
find labels indicating that these houses were
monuments because there are maps where all of the monument canal houses in Amsterdam are positioned. And if I would be able to
indicate something unique, for example, four houses
next to each other, being a monument, then I would be able to look at the map to see
where on the Amsterdam canals, four houses would be
connected to each other and put myself there with Street View to confirm the location. Also the picture wasn't
clear enough to zoom in that far to see if I can
find the house label. So this is probably a house number or a label indicating
which family lives there, because I know the monument
tags are either square or rectangle, so they would
never be oval or round. One of the other things that
caught my eye on this photo was the orange canal
cruise boat in the canal. If you zoom in, you would
only see the name Brussels because most of the canal
cruises give their boats names. So I Googled to see what
kind of canal organization would have a boat named Brussels. And it was the Lovers company. So the Lovers company
uses orange tinted boats in the canals in Amsterdam. And of course on their
website, you can buy tickets to all of the various routes
they take to the canals. My best guess was let me
take a look at the routes to see where they were going and then follow the routes on Street View in order for me to see if
it passes the same direction or if it passes the same area. I also knew that from the
area we were focusing on, like from where the person was standing, where the picture was
taken, he was on a bridge. So it would be a canal. There would be like a crossing of canals because you can see a bridge
slightly in the bottom right. You can see the like arch shaped. So there's probably a canal
going underneath as well. So these were my two points to go at. And actually as I was
cruising in Google Street View to view where the boat was going. And one of the first
points where I started because from Central Station,
which is in the center, all of the canals are laid
as an onion around them. And often the canal tours
start from Central Station and then go into the canal. So I started at the beginning point of where they start as well. And I did it to this one too. And I immediately landed
on exactly this location just by pure luck. It could have taken me far longer if I would've taken the other route, which the canal cruises follow as well. I just was very lucky. So this took me over a
half an hour to geolocate. - Is that good or is that
like quite a long time? - No, it's quite good. I already knew where this was, right? Because I had a hunch this
would either be an Amsterdam or in Utrecht where we have similar view, although the canal houses
are not built that tall. So the person who took the
photo is probably standing a little bit more back towards the left, but we can see the light pole here and we can see the same
shaped canal houses as well. And as we can see, there
was some work on the canals. We see some gates that were placed there. So it's not exactly the same time and date as the picture was taken. But we can see that the same shape of houses are still there. So even though this was a
very familiar area for me, the moment I was standing
here, I was like, "I crossed it a million times when I was cycling in Amsterdam." But you never pay any attention
to the shape of the houses for some reason, probably. So I didn't recognize it instantly, like, "This is the exact location." So it took me some time to figure it out. And actually the canal
cruise was giving it away because it gave me a good direction on which canals it could have been. Otherwise it could be like anywhere. And I tried to figure out how
many canal houses there are, but like there must be over
2,000 or something of them. ( laughs) Even though it was
a very familiar location, I had to pivot onto other points in order to find the exact location. - So do you think it's
easier to do like cities or is it easier to do like countryside? - I would say cities generally, because there is usually
far more information in a urban setting than
you get in a rural setting. So in a city you get
cars with license plates to identify countries,
you get street signs, you get road names,
you get business names, phone numbers on billboards. They're all pivot points,
they are all research points. You get dumps in the
middle of the countryside, just because trees, fields, hills are much less distinctive. And so yeah, urban environments
definitely are easier. - For example, the BBC,
especially the "BBC African Eye", they geolocated a video of a couple of military men shooting
women in what seems to be like a desert area. And by looking at the mountain ridges, by looking at the kind of
crops that was growing there, by looking at the state
some of the trees were in to indicate what kind of season it was. And they spent tons of
times on Google Maps, just like Steven did with Loss Ness to look at the mountain range, to see if we would match with
what they saw in the video. And it took them over two
weeks and it only helped them because they received an anonymous tip that it would be in a specific area. And then they narrowed down the search and they were able to
match the mountain ranges. But a countryside area, especially when there aren't
a lot of pivot points, can take you a lot of time. - You both use this term
pivot point quite a bit. What does that mean? - These are aspects within the photo that you could help to locate. So for example, looking at the
light pole here in Amsterdam, I would go to see, Hey, what kind of light poles
are used in the Netherlands? Are these specific for a specific area or are they only used in,
for example, Amsterdam? Or looking at the number
of the parking zone, that would be something
for me to take and go to, for example, Google, to see if I can find a
location on the world where they use that
specific kind of thing. So that would be a pivot point. - The more you can find
often the easier it is. There's a really good exercise actually that helps train people who
are new to this sort of thing. I was speaking to someone
in the British army, who trained intelligence officers. And this is what was one of his exercises that he recommended. So he would give his
students a photo and say, "Right, I want 20 bits of
information from that image. And if it takes you all
day, it takes you all day." And you'll be amazed if you stop and forced yourself to look at a photo, you can pull out details, which might not seem very relevant, but they have a springboard to go and research something else. So if someone's asking me to
geolocate this image in front of me here, say, "Right, I
want 20 bits of information." So you might say, "Okay,
there's a street sign. There's this yellow sign with a B in it. What does that mean? There are yellow license plates. What countries have that? There is a 15 kilometers an hour sign. So which countries have
that as a speed limit? Where do I find lampposts
that look like this? Where are there lots of bicycles? What does this writing here say?" In a photo like this, it's
very rich in information, but if you can learn that technique, then you can apply it to
photos where there doesn't seem to be so much information. You set yourself a target of, "I'm gonna write down
10 things in this photo or 20 things in this photo." You actually train yourself to be able to extract information that isn't always
immediately obvious to you. Because sometimes you can be working with not very much information at all. So every little bit you
get can be really helpful. - So I just wanna say
thanks to both of you, Steven and Lizette for sharing. You've shown us a lot of
really cool ways to find stuff. Just for the audience, if you
want us to do more of these, please put in the comments below. If you want me to like
travel to interesting places and use that as photos, let me know. Thanks to both of you. I really appreciate you sharing this and hopefully we can get you back soon. - You're welcome. - Yeah, it's been great, thank you. (upbeat music)
