The article will help beginners to recognize signs of explosive objects in images found on social networks. Learn how to identify unsafe content using the tools and techniques used by content verification experts. You’ll gain valuable information on photo analysis, learn useful resources for munitions identification, and expand your OSINT knowledge.
Explosive munitions (EDs) pose a serious threat. You should not approach them unless you have special training. When posting information about suspected explosives on the Internet, it is important to avoid encouraging anyone to interact with them. If a potential WB is identified, the police or other competent authorities should be contacted immediately.
Landmines, rockets, bombs, grenades, and other explosive devices that failed to detonate during hostilities can remain active and dangerous for decades after conflicts end. These unexploded ordnance can detonate at any time, causing serious injury or death to civilians, even years after the end of the war.
According to data obtained in 2023 from Landmine Monitor, more than 4,710 people were injured or killed by landmines and explosive remnants of war. In Ukraine, between February 2022 and May 2023, the HALO Trust recorded 855 civilian casualties or deaths in 550 mine-related incidents.
The proliferation of images of conflict on social media has provided researchers and civil society organizations with a valuable resource for identifying explosive objects, particularly ELVs. At the same time, it increased the risk of misinformation and disorientation about the use of weapons. Improper identification of WBs can spread false narratives, create false accusations during hostilities, and create dangerous situations for deminers.
Once you’ve established that an image is authentic and that the subject it depicts has not been previously identified, closer examination can often provide clues as to what the subject is.
Larger ammunition usually has markings or stamps indicating the specific model type, date of manufacture, or production lot number. If the items themselves or related objects nearby, such as ammo crates, have visible inscriptions, this can make it easier to identify the model.
For example, on weapons manufactured in NATO countries, the CAGE code (Commercial and Government Code of the Organization) is often found – a unique five-digit number assigned to suppliers of defense products for government or military institutions. This code, available in the Defense Materiel Administration’s online database, can be used to trace the origin of explosive devices. Such codes have already been used to analyze the origin of ammunition in conflict zones, in particular in Syria and Yemen.
Due to the huge variety of existing EO models, there is no centralized resource for identifying the meaning of labels and markings, even if they are clearly printed on the object. However, there are various guides that focus on specific types of EOs or specific origins. For example, the Collaborative ORDnance Data Repository (CORD) has a database of more than 5,000 records related to the detection of landmines and other explosives. Another example is Bomb Techs Without Border’s Guide to Basic Identification of Munitions in Ukraine, which covers all known munitions used during the Russian invasion of Ukraine.
The color of an explosive object is also an important characteristic for its identification. Different colors can indicate the type of explosive inside or the purpose of the object. For example, training or practice ammunition in NATO countries usually has a blue body color with white lettering. Such markings allow you to quickly distinguish between warheads and training models, reducing the risk of accidental dangerous incidents.
The nonprofit GlobalSecurity.org has a list of color codes used by the former Soviet Union, the United States, and other NATO countries.There is also a helpful guide on how to identify bullets, including explosive rounds, by the color markings on their tips.
If the item has no immediate identification marks, label, stamp or color, the next step is its shape and size.
You can view the photo below for definitions of the types of EOs mentioned earlier, their general shapes, and sample images.
The given descriptions are not exhaustive and cover only the most common models in each category. Explosive objects can have a wide variety of shapes, sizes and designs. The proposed decision tree below can be a useful tool in the identification process, providing basic guidelines for determining the type and characteristics of ammunition.
Depending on the quality and additional context in the image, it is sometimes possible to measure the approximate dimensions of the object.
If the type or model of EO shown in the image is still ambiguous, there are a few clues that can help further narrow the scope of the study and provide useful context.
Some types of weapons are only used in certain regions, so the geolocation of an image can be an important clue for investigators in determining the type of explosive device.
For example, Russia did not conduct large-scale ground operations in western Ukraine at the time of its invasion, so images from the region likely depict advanced aerial assets such as long-range cruise missiles or aerial bombs.
Cluster munitions, which scatter numerous submunitions over a large area upon detonation, are known for their danger to the civilian population. They were used in 41 countries and five territories after World War II. These munitions are banned by an international treaty ratified by more than 100 countries because of the significant threat they pose to civilians. This restriction makes it possible to narrow the circle of countries that can be a source of such ammunition.
The timing of the incident is an important factor in determining the origin of the weapon. For example, if the event in Ukraine took place before 2016, the probability of using Western weapons is much lower, since Western support for Ukraine was limited before that period.
The timing of the event also helps narrow down the search for relevant arms transfers that may have occurred prior to the incident. The Stockholm International Peace Research Institute (SIPRI) provides access to a database of major conventional arms transfers from 1950 to the most recent complete calendar year, allowing the use of incident date information to narrow down the range of possible munition types.
Also, certain weapon systems can be excluded if the image is dated to a time when that weapon was not yet developed or deployed. For example, the Russian PTKM-1R anti-personnel mine was presented to the public in 2021 and was first used in Ukraine in 2022. So, if the image is taken before these dates, this model is probably not present. However, sometimes new types of weapons can be used in a limited manner even before their official disclosure.
Researchers also use “chronolocation” techniques to determine when a photo or video was taken. This process involves analyzing the length of shadows or comparing landmarks in the images to satellite images taken at different times to determine whether visible features of the landscape have changed over time.
Images and videos on social media rarely focus on a single object, so if they contain multiple objects, they can be useful clues to the identification of explosive ordnance (OBD). The imagery may include other munition-related objects such as casings, fuzes, debris, shipping containers, or special mechanisms that may provide additional information about the type and model of the munition near the target object. For example, ammunition containers are often marked to allow quick identification of their contents. In addition, some types of mines have plastic safety covers that are removed before use, and which are often left nearby, adding valuable identification clues.
After getting acquainted with the methods of identifying explosive objects based on images from social networks, we will consider a practical example of the application of these steps.
Let’s imagine that we have a picture from a social network in front of us, which shows a suspicious explosive object in the Budyoniv district of Donetsk, in the east of Ukraine. This footage was published on the popular channel “Typichny Donetsk” on the Russian social network “VKontakte”, which often covers the fighting in this region. By analyzing the image, you can try to determine the type of ordnance, paying attention to accompanying objects that can help establish the origin and model of the explosive object.
Are there any inscriptions on the object?The answer is no. The item appears to be damaged and rusted and unfortunately any identifying marks are either on the side in the ground or have been obliterated by use or weather.
What colors are represented on the item?The object is light green in color but also heavily rusted to the right of the image. Given the object’s deteriorated condition, it is difficult to identify its original color and any color streaks that may have been present.
What is the shape and size of the object? Does it have any unique features?The object has an oblong shape with easily recognizable fins and spikes to the left of the object. To the right of the widest part of the case there are circular marks which can also be recognized as the obturator rings used to seal the gases for the artillery, despite the heavy rust.
Referring to our decision tree, we can identify it as a type of mortar ammunition. This can be confirmed by checking the Open Source Munitions Portal for other confirmed mortar images. However, we cannot narrow it down to a specific model without additional sources, as the object shown is heavily corroded and has no visible stamps or inscriptions.
This guide is dedicated to the identification of explosive ordnance (ED) in social media images. However, due to the variety and complexity of ammunition types, accurate identification may require referral to specialized resources. We have collected the most useful of them in a special repository.
The repository is a dynamic document that is regularly updated with new resources after they are published. Each of these resources is marked with an appropriate category, as well as information about the origin or region of use of the ammunition it describes. A brief description of each source has been added for ease of access. Most resources are free and open to users, but some require registration or offer paid access.
In addition to human object identification, computer vision and machine learning tools are increasingly being integrated for WB recognition. Although these technologies are still in development, startups like VFRAME and Tech4Tracing are actively working on it. In the future, this will help researchers identify explosive objects not only in static images on social networks, but also in real time from drone images.
If you’re an iPhone user, you’re probably familiar with image names like “IMG_8928.jpg”. This format includes the prefix “IMG” followed by an ascending number for each image. However, this naming system is quite general and does not provide much information that would be useful for open source researchers.
But if you’re using an Android phone, you might be more familiar with the filenames that contain the date and time the image was taken:
PXL_20240830_150806479.jpg
IMG20240830150806.jpg
IMG_20240830_150806.jpg
20240830_150806.jpg
These file names can be useful for researchers seeking to chronology images or establish the order of events at a particular location. File names can also give clues about the type of device on which the image was taken. For example, if the file name starts with “PXL”, this indicates that the photo was taken on a Google Pixel device, and the date format may indicate that the image was taken on August 30, 2024.
However, not all file names displayed correspond to the original image names. Some files may be renamed based on the date they were uploaded, such as “2024-08-30.jpg”. It’s important to be careful with these names, as they likely indicate the date the image was uploaded rather than the date the image was taken.
Timestamps are usually the most common information to be gleaned from image file names, but this tool can also provide other clues, especially if the image was renamed before uploading to Google Maps.
While investigating alleged links between a company called CasinoMentor and an AI art website called OpenDream, a Google Maps listing for CasinoMentor in Malta was found that contained the following photo uploaded by the company.
A reverse image search revealed that the photo was taken in Vietnam, where we suspected OpenDream’s founders were also located.
To better understand the context of this photo, we paid attention to the file name. It turned out to be “CMTeam.jpg”, which suggests that the image is of the CasinoMentor team. Thanks to geolocation, we were able to confirm that the team was indeed in Vietnam, despite the company’s official address being in Malta.
1243 
1063 
1289 
