Photo authenticity verification

Verifying a photo's authenticity based on evidential context is a transparent and reproducible mechanism that takes into account shooting conditions, time, location, and other related data available on the user's device.

Online verification of shooting context by identifier

To start the verification, it is enough to enter the unique identifier of the photo (PUBLIC UID) in the input field and start the verification. After that, the system will display a set of data generated at the moment of shooting and during subsequent processing. This data allows you to understand when, under what conditions, and within what process the photograph was taken. Important: the verification does not analyze the image visually. It shows the evidentiary context associated with the photo creation process.

Public identifier (PUBLIC UID)

What data is displayed and how it helps verify the context

PUBLIC UID — public photo identifier

A unique identifier used to perform the verification. It unambiguously links the photograph to a record in the system and prevents the results from being replaced by another photo.

client_captured_at — shooting time on the device

Reflects the moment of shooting according to the user's device time. This field allows:
- determining the chronology of events;
- matching the photo with the declared work execution time;
- detecting attempts at backdating.

is_verified — photo integrity status

Shows whether the photo has been altered after shooting.

- true — the image has not been changed after creation in the app;
- false — the photo has been edited or re-saved after shooting.

Important: the false value does not mean the context is unreliable, but indicates that the visual integrity of the image has been compromised.

timezone — shooting time zone

Displays the time zone in which the shooting time was recorded. This allows correct interpretation of time data and eliminates errors associated with different time zones.

lat and lon — shooting coordinates

Latitude and longitude recorded at the moment of shooting. Used for:
- verification of spatial context;
- matching the photo with an object or inspection area;
- analysis of the logical consistency of routes and actions.

gps_accuracy — coordinate determination accuracy

Shows the location determination error in meters. Allows assessing the reliability of coordinates and understanding the conditions under which they were obtained.

address — shooting address

Displayed as a string address determined at the moment of shooting. Used for human-readable verification and matching with the declared work location.

model — device model

Allows understanding on which device the shooting was performed. This is important for:
- analysis of data acquisition conditions;
- detection of anomalies;
- confirmation of the use of a real mobile device.

platform — device operating system

Specifies the OS on which the application was running during shooting (e.g., Android or iOS). Helps correctly interpret the specifics of data collection.

app_version — application version

Records the version of the application in which the photo was taken. This allows accounting for changes in data recording logic between versions.

created_at — record creation

The moment of creating a record in the system. Used to verify the consistency of server time and client data.

updated_at — record update

Shows whether the record has been updated after creation. This helps understand if metadata changes have occurred and at what point.

The obtained fields collectively form an evidentiary context of the shooting, which allows separating the verified facts from interpretations and assumptions.

What does "photo authenticity verification" mean

Photo authenticity verification is not an attempt to "guess" whether a photo is genuine or not, nor an analysis of image pixels. Within the INSPECTOR project, authenticity is understood as the credibility of the shooting context: confirmation of when, where, under what conditions and under what circumstances the photo was taken, as well as which facts can be confirmed and which – fundamentally cannot.

The app deliberately separates the verification of the photo's context from:
- image analysis using AI;
- searching for traces of pixel editing;
- subjective assessment of the image's content.

The app's task is to provide the user with verifiable and reproducible evidence, not interpretations.

What can be checked on a photo

1. Shooting context

The context of a photo is the set of conditions under which the shot was taken. Within the photo context verification, the following can be confirmed:

the time of image creation;
the sequence of events before and after shooting;
the photo's connection to a specific report, task, or event;
the continuity of the recording chain (when and by whom actions were performed);
the photo's compliance with the declared shooting purpose.

Important: it is not about the "truth of the image," but about the truth of statements about it.

2. Photo shooting conditions

We can confirm:

that the photo was taken by the user and not uploaded externally;
that the shooting took place within a recorded scenario;
that the photo was obtained at a specific moment and not backdated;
that the user's actions during shooting complied with the defined process.

This is especially important for:

report photos;
photo documentation of completed works;
inspections, checks, audits;
documentation of object conditions.

3. Related data (evidentiary context)

The photo is considered not in isolation, but as part of a data set. Subject to verification are:

user actions before and after shooting;
the logic of transitions between process steps;
the fact of saving, transmitting, and using the image;
the immutability of the recorded context after the photo creation.

It is precisely this data layer that forms the photo's evidentiary context.

What cannot be verified from a photograph

1. Image Content

We do not claim and do not verify:

that the depicted object is 'real';
that the event in the photo happened exactly as the viewer interprets it;
that there are no staged elements in the photograph;
that the image cannot be visually imitated.

Any visual interpretation always remains subjective.

2. Absence of Pixel Editing

Verifying a photo's authenticity is not equal to verifying for editing. We do not claim:

that the image has not been processed with graphics editors;
that there are no traces of correction in the image;
that the photo is 'original' in the technical sense.

Even a fully edited image can have a reliable shooting context if that context was captured correctly.

3. Intentions and Interpretations

A photograph does not prove:

the motives for actions;
the causes of what happened;
the legal rightness of the parties;
the assessment of the event's consequences.

Context confirms facts, but does not replace conclusions.

How we prove the authenticity of the shooting context

Principle

The authenticity of a photo's context is confirmed not by the image itself, but by the process of its creation. Key principle: > if the process is recorded, reproducible, and consistent — the context can be considered provable.

Stages of Forming Evidential Context

Scenario Recording — the photo is taken not randomly, but within a specific action.
Sequence Control — the user's actions are recorded in a logical chain.
Link to the Inspection Object — the photo is tied to a specific task, object, or report.
Immutability After Creation — the context cannot be rewritten retroactively.
Reproducibility — an independent party can understand how and under what conditions the photo was taken.

How Context Verification Differs from AI Analysis

Context Verification	Image Analysis
Checks conditions	Analyzes pixels
Process-based	Probability-based
Reproducible	Often non-deterministic
Explainable	Model-dependent
Suitable for reports	Suitable for filtering

Context verification does not replace AI, but solves a different problem.

Where is this applied in practice

Inspections and Checks

photo reports on the condition of objects;
control of work execution;
technical inspections;
inspector checks.

Context is more important than the appearance of the image.

Business and Contractors

confirmation of service completion;
reports for clients;
resolution of disputable situations;
control of remote performers.

Journalism and Research

confirmation of photo origin;
verification of material acquisition conditions;
distinction between facts and interpretations.

Legal and Expert Environment

preliminary assessment of photographic evidence;
analysis of accompanying conditions;
exclusion of context substitution.

Important: context verification does not replace expertise, but increases transparency.

Method Limitations

We consciously indicate the limitations:

context is not equal to truth;
a photograph does not prove the entire event;
any conclusions require interpretation;
the method is not intended for visual expertise.

Openly stating the boundaries increases trust in the results.

Summary

Verifying a photo's authenticity through context is a way to:

separate facts from assumptions;
confirm shooting conditions and process;
record evidentiary context;
honestly show what can and cannot be proven.

It is precisely transparency and reproducibility that make such verification useful for users, businesses, and professional environments.

Frequently Asked Questions about Photo Authenticity Verification

Is it possible to verify a photo's authenticity without image analysis?

Yes. Within contextual verification, it is not the image as a set of pixels that is checked, but the conditions and process of its creation: when, in what context, and under what user actions the photo was taken.

Do you check if a photo has been edited?

No. Authenticity verification through context is not intended to detect image editing. Even an edited photo can have a reliable shooting context if the process was correctly recorded.

How does context verification differ from AI analysis of a photograph?

AI analysis works with visual content and probabilities. Context verification relies on a recorded process, a sequence of actions, and reproducible data, making it explainable and verifiable.

Is it possible to prove the shooting location from a photograph?

Context verification does not claim to confirm the geographical location as a fact. Only the conditions and actions that accompanied the shooting within the recorded scenario are confirmed.

Is such verification suitable for legal and expert tasks?

Contextual verification does not replace expertise, but can be used as an additional source of transparent and reproducible data when analyzing photographic evidence.

Can shooting context be forged?

The context model is designed to exclude the alteration or retrospective rewriting of shooting conditions. However, the method honestly states its limitations and does not claim absolute truth.

For which areas is context-based authenticity verification particularly useful?

The method is applied in inspections, technical examinations, business reporting, journalism, research, and preliminary legal assessment of photographic materials.

Does context verification prove that the event actually occurred?

No. The verification confirms the shooting conditions and process, but does not interpret the image content and does not replace conclusions about the event itself.