PeakVisor
Dual window views for any global location: (1) a 2-D map & (2) a 3-D rendered terrain model, with photo fitting, shade/slope mapping, sun trails & weather data. In active development for OS research.
Last updated
Was this helpful?
Dual window views for any global location: (1) a 2-D map & (2) a 3-D rendered terrain model, with photo fitting, shade/slope mapping, sun trails & weather data. In active development for OS research.
Last updated
Was this helpful?
https://peakvisor.com
PeakVisor is a multi-function OSINV tool for examining high-accuracy geographic data in the form of 2D and 3D maps and rendered terrain models of near-photographic quality. This geographic data is also overlaid with detailed sun/moon, slope, shadow, elevation and, where available, weather information. PeakVisor is a leading tool for dramatic terrains, e.g. mountains, coastlines, canyons...
Beware that PeakVisor comes in 2 versions: the Web-Based App, used for Open Source research, described here, and the Smartphone App, used for outdoor navigation, which has significantly different functionality. Online resources can mislead as they often don't specify which version they cover.
Location co-ordinates (latitude and longitude)
Altitudes (height above sea level)
Viewpoint orientation angles (pitch, yaw, and roll)
Field of view/focal length angles (degrees/mm)
Dates (dd/mmyyyy)
Times (hh:mm)
Sun/moon positions (azimuth and elevation angles in degrees)
Source imagery
Geographic paths ([lat1, long1],[lat2, long2], [lat3, long3]...)
Pixel in a 3D panorama (on-screen pixel select by user)
PeakVisor will not search for a matching location for your source imagery, you must provide a test location, i.e. a "guess" in order to use PhotoFit for confirmation. Obviously you can repeat the process, testing different locations, until hopefully a match is found.
IFF the PeakVisor terrain outline fits very well over the same characteristic features in the photograph, then this is very strong evidence of an accurate geolocation. The more points of inflexion* there are in the contours being fitted, the stronger the confirmation of location. The example above is a very visually obvious geolocation, chosen to demonstrate the principle. The source image (a photograph of the Matterhorn mountain) has been uploaded to PeakVisor in "Explorer" view, and the user is moving the red silhouette of the model (which PeakVisor generated at the test location) across the photograph to try to adjust it to fit. * Point of inflexion = change in direction
PhotoFit Adjustment Can Be Time-Consuming: The user has a selection of variables which can be adjusted to obtain an image match: viewpoint location, viewpoint angles of pitch, yaw and roll, viewpoint field of view/focal length.
Remember that the camera may have moved or zoomed between images if your source imagery is a 'spliced' panorama, so the goodness of fit may change between panels.
Visually, changing Field of View and changing Viewpoint Location can appear confusingly similar. They are not the same thing: the first is like zooming a telescope to change the view whilst remaining stationary, the second is moving closer or further away to change the view. Generally, changing Viewpoint Location is the more important means of obtaining a good geolocation PhotoFit.
Most cameras are set to around 60 degree Field of View. Exceptions which would justify changing the Field of View to obtain a fit are the use of zoom, security camera imagery, which often uses high angle 'fish eye' lenses.
because no images have been taken, e.g. in a very remote area OR
because they are unavailable for geopolitical reasons, e.g. war, government intervention.
Use PeakVisor in conjunction with other tools like Google Earth Pro and satellite imagery to obtain the fullest understanding of Test terrain and environment for which geolocation is needed. Exploit the differences of these tools, e.g. Google tools are quick to cover terrain in real time, strong on time-stamped historical data, strong on local photographic imagery, weak on exact photographic location, unclear on photographic aspect. PeakVisor is weak on cityscapes, sparse on local photographic imagery, slower to switch viewpoint, but strong on orientation, clear on numeric viewshed data and strong on distance measurement in 3D panoramas.
PeakVisor provides an elevation data download feature for a maximum ground surface rectangle area equating to 0.5 degrees square of latitude and longitude. The user can define the location and shape of the download rectangle data and save it as a *.png file in the Data tab of the PROFILE page.
The developers have reports of problem using the Teleport function from users and look to resolve them by July 24 during an major upgrade.
The Distance measurement feature can be used as a rapid way to determine lines of sight, i.e. using PeakVisor output we can say "If this image was taken at location L1, then the following features would be visible because there is a line of sight to their locations from L1 (and vice versa)." Where the horizon is chosen as line of sight, this can be a very useful way to determine what should and should not be visible from a given location.
The PeakVisor view above could be used to test whether the location, time and date of geolocation source imagery were consistent and feasible, e.g. at the test location, could the sunset still be visible from the viewpoint at the time and date claimed?
There are also view types for Skiers and Hikers, each of which has functionality useful for Open Source research:
Slopes: Map colour shading to indicate degree of slope of the terrain, from white for flat, through yellow for mild to orange and red for steep.
Very useful for doing wide-area searches to match major slope terrain features and thus reduce search space in a geolocation.
Shadows: An overlay view generated around a user-input date and time, where the shadows cast by the landscape are rendered in detail.
Move the time slider to watch the major shadows cast by terrain features move across the landscape at the date, time and place you choose. Very powerful for confirmation of geolocation imagery containing terrain shadows, e.g. how far does the shadow of that mountain/canyon side reach at a certain time of day and year?
It is easy to use PeakVisor's basic features, but the most useful ones take some time to master.
Similarly, the Sun and Moon trails feature takes time to master for chronolocation purposes, and the use of the centre screen crosshairs (SHIFT + P) is needed for accuracy.
An email login is needed for some functions like elevation data download, location and path saving. This can be created using the SIGN IN feature in the centre of the landing page - see below.
Urban Environments Rarely Usefully Rendered: PeakVisor Source Data is greater than 30m/pixel in many areas, but features smaller than 30m cannot be assumed to be visible on PeakVisor Explorer View rendered terrains. Thus **Urban environments** will rarely be usefully rendered with respect to bulidings, but useful information can be gleaned about urban lanscapes and lines of sight.
Resizing Windows: PeakVisor 3D and 3D Explorer windows can switch zoom scale when Teleporting between viewpoint locations in the model, which means users often need to resize and adjust scale during their work.
Buffering Effect at Capacity Limit: PeakVisor operates at the limit of capacity when the user continuously adjusts viewpoint in 3D Explorer View, because it must constantly re-render the imagery. If buffering limits are reached, the system will 'boomerang' back to a previous location. The way to work around this is to move the viewpoint in increments.
PeakVisor can and has been used in geolocating imagery from remote areas which claims to depict mistreatment of indigenous minorities. As such, it can be used support to work which seeks to protect human rights, indigenous peoples, minorities and the environment through accurate geolocation. The type of exact (within 30m, sometimes 10m in some areas) geolocation which tools like PeakVisor and satellite imagery (sometimes used in conjunction) can deliver can indirectly reveal the identify of people providing source information, which could put them/their contacts at risk of reprisal. Source information providers may or may not be aware of this risk, so there is a responsibility on the Open Source researcher, who may have more information for accurate risk assessment, to evaluate that risk on their behalf. Many information providers willingly undertake the risk of reprisal for the benefit of drawing public/global attention to situations and events they hope will be changed or acted upon as a result of exposure.
Sophie Tedling, Bellingcat Website
Article on an OS Investigation using PeakVisor
Sophie Tedling, YouTube, Bellingcat Channel
Video Webinars:
Sophie Tedling, YouTube, Bellingcat Channel
Video Webinars:
X/Twitter
X/(Twitter) Account
The PeakVisor website help does not refer specifically to the Web Version, and so describes some major features which are only available on Smartphones.
Routes Software SRL, VIA CAVOUR 2 LOMAZZO, COMO, 22074 ITALY
Sophie Tedling.
The web-based version has been developed specifically for Open Source research since 2022 via the POPSI project (on X/Twitter: ). It allows users to input numerical values to describe 3D geographical viewpoints in terms of latitude, longitude, altitude, pitch, yaw, roll and horizontal angle of view. It superimposes maps over geolocation source imagery to fix location by matching map to historic recorded views. This article covers the web-based version for Open Source research. This is best used with a large screen and an exact pointing device like a mouse. A free login is required to access all features.
PeakVisor's most widely-used and praised feature to date is that it provides a geolocation function - see below.
The PhotoFit feature is particularly useful where local "Streetview" imagery is not available, e.g. in the usual sources like , , , etc., either
(3D rendered terrain model displayed with 2D Map View)
As well as , which always displays a 3D rendered terrain model of a given location next to , PeakVisor provides a number of other view types for its topological data:, , , and . These are accessible from the Drop Down menu under Explorer in the Toolbar - see below:
The PhotoFit feature often requires a good deal of adjustment to obtain a good fit, and some understanding of visual field of view and focal length. and note that many different small adjustments are needed to get the skyline fit.
Performance: PeakVisor Web version for OSINV is a development project, so performance can be volatile whilst upgrading. See
Five USP PeakVisor Features: 30m Global Accuracy Data, Contour Mapping for Lines of Sight, Sun Trails for Chronolocation, Distance Calculation within Virtual Terrain, Photofit Image Match Validation
PeakVisor Map Interface: Explore Mountains Map Page
Seven Underlying Parameters Describe the PV Panorama
How to Input a Location
How to Adjust Altitude and Read Contours in the 2D Map
How to Adjust Viewing Angles
How to Adjust Field of View and What It Means
How to Move from an Existing Location
How to Measure Distance to Another Location
How to find a PhotoFit
Sun and Moon Trails and the Shadow Method: Suncalc
Depth and Field of View
Field of View, Focal Length and HFOV: Chasing the Perfect Photofit
