Thursday, July 13, 2023

New technology sees through Russian attempt to hide ships from Ukraine

Analysis of a radar satellite image of Russian warships, some with the new camouflage, in Sevastopol, Crimea.
The artificial intelligence (AI) ship detection algorithms used by Satim Inc. can clearly differentiate different classes of ship even with the camouflage applied.
Satellite image from Umbra Space.

From NavalNews by H I Sutton

Ukraine and NATO will be closely monitoring Russian naval activity in the Black Sea.
Russia has however tried to make this more difficult by devising a unique new camouflage scheme.
Modern technology can however help unmask these deceptively camouflaged warships.

Russia is attempting to hide important warships in the Black Sea by painting them in a new deceptive camouflage scheme.
The ruse makes it difficult to judge their length and shape, and so identify them in satellite imagery.
This could make it more difficult for Ukrainian planners and, potentially, even thwart an attack.

Defense analysts have been exploring how effective it is.
Initial checks in some satellite images confirmed the viability of the camouflage.

However, Imagery analysis company Satim Inc. has shared a way in which artificial intelligence (AI) can help see through the camouflage.
Their AI can identify classes of ships in radar satellite imagery, which is very difficult to do as a human.
This approach is unaffected by the camouflage.

This deceptive camouflage is unique to Russia in modern history and is distinct from the more famous Dazzle camouflage.
Dazzle attempts to confuse an observer so that they misjudge distance, speed and heading.
Here the aims are quite different.

The New Camouflage Makes Some Sense

Since Naval News first broke the story of this camouflage June 22, several more Russian warships have been observed with this scheme.
The flagship of the Black Sea Fleet, the Admiral Grigorovich-class frigate Admiral Makarov, has had a more extreme camouflage applied.
This includes ‘blacking out’ the helipad and the bow of the ship.
Then all three of the Buyan-M class missile corvettes, the minesweeper Ivan Golubets and the anti-submarine corvette Muromets, have the camo.

Guide to Russian Navy warships with new deceptive camouflage

There are indications that the camouflage is intended to hide the overall length and form of the ship when viewed from the surface, such as by a maritime drone.
However the larger factor is likely from above.
Here the camouflage is intended to make it harder to identify the ship in satellite imagery.
This is significant as we can reasonably expect that both Ukraine and NATO are using satellites to monitor Russian fleet movements.

The camouflage isn’t likely to fool an experienced analyst if they have time to study the image.
But it might not have to still be useful to the Russians.
How well the camouflage works will depend a lot on the light conditions, the satellite being used and the time available to analyze it.
Overall we can expect it to slow down the work of analysts and increase uncertainty.
This will be especially felt while creating a picture of overall fleet disposition.

Another factor is the use of artificial intelligence.
Countries can leverage AI to speed up and scale-up image analysis.
This AI may try to determine the length and shape of a ship to identify it.
So it may actually be more vulnerable to these types of camouflage than a human.

Very interesting for anyone doing OSINT, impressive ship detection / identification algorithms from
courtesy of Richard Culshaw
Seeing Through The Camouflage

Known as SAR (Synthetic Aperture Radar), radar satellites are not new.
However it is only recently that commercial satellites have been able to provide high-resolution SAR imagery.
Unlike regular satellite imagery which passively interpret what they see, SAR radars produce their own energy and then records the amount of that energy reflected back to whatever is on the earth.
The imagery can be beautiful, but it can be difficult to interpret as a person.
Ships appear blurred and the parts which reflect the energy may not be intuitive, masking its form.
Experienced analysts may be up to the challenge, but there less certainty than typical optical satellite imagery.
Certainly, differentiating two distinct classes of warship can be very difficult.

Satim Inc. have an answer.
Their artificial intelligence is trained to recognize distinct classes of warship and submarine.
Their system has a level of accuracy over 90 percent.
Satim have been able to demonstrate to Naval News that their technology can identify the relevant classes of ship in the Russian base at Sevastopol, Crimea.
Using a SAR image taken by an Umbra Space-owned satellite, it clearly identifies ships which are known via other sources to have been camouflaged.

Another AI analysis of Russian Navy ships in Sevastopol by Satim Inc.
The Buyan-M highlighted has the new camouflage.
Note how the AI can differentiate the different classes of ship even though they look similar to the human eye.
Satellite image from Umbra Space.


Combined with other satellite imagery and other forms of open source intelligence (OSINT), the AI interpretation of SAR imagery can help unmask the camouflaged ships.
SAR has additional advantages that it works at night and can see through clouds.
Thus it increases the persistence of satellite coverage.

Russia will likely continue to experiment with efforts to deceive satellite monitoring.
Last year, the Russian Navy has attempted to disrupt imaging of ports and bridges by placing barges with radar reflectors nearby.
This is generally agreed to have not worked, and the later abandoned the practice.

There are already indications that the first warship reported with the new camouflage, the Admiral Essen, has had some or all of it removed.

However there is not yet any sign that it is being discontinued elsewhere.
And new ships may soon be identified with the camo.
Analysts will chase developments and adapt; the game goes on.

Links :

No comments:

Post a Comment