Face recognition system
True Detection Security
About the System
Based on artificial intelligence (AI) and biometric identification our system provides recognition of:
Faces
Silhouettes
Objects
Vehicles
License plates
Challenges
Also, a serious risk factor is the lack of timely and sometimes immediate response to a possible security threat.
Unfortunately, nowadays, security threats are still exist all over the world. Namely, such incidents as:
Stay of illegal and undesirable persons in the country.
Minor and domestic crimes;
Vandalism and criminal behavior;
Car theft;
Street theft and robbery;
Proposed Solution
Set up an automatic system for alerting security personnel about the presence of a threat.
Create a risk model to assess the potential security threat.
Create end-to-end identification of citizens and transport using a recognition system.
System Features
Human faces and silhouettes recognition
Creation of watch lists (face image + any metadata) for further monitoring and recognition of these faces.
Real-time face recognition based on watch lists using any surveillance cameras and video streams (video from YouTube, Shorts, Tik-Tok, etc.).
Analytics of objects of observation
Construction of social graphs.
Displaying places, meetings, names of events attended by the surveillance object*
Displaying the watch lists where the surveillance object* was recognized.
Displaying the total number of surveillance objects* (recognized and unrecognized people, with the possibility of sorting) after each event in a report.
Additionally
Searching for people by photo or video (if this person was previously caught by any camera or was in the database).
Customizing notifications (for example, if the system recognizes the person, it can send a personal notification about it).
Ability to create users in the system with different access rights.
Searching for fake news.
Recognition of any symbols on the video (chevrons, swastika, flags, etc.).
  • the number of visits to a particular address.
  • at what address the person was seen in a given period of time;
Searching for people on the map:
Face Recognition
Identification of individuals (groups of individuals) associated with extremism, drug trafficking. In a real-time mode, the system recognizes faces on video, compares them with «web users» or «watch lists» pre-loaded into the system.
Revealing Additional features
Gender, age, emotion, spectacles, beard, mask etc
Identification and verification
  • Searching people in tourists' data base
  • Searching in security data base;
  • Searching in criminal data base
Individual detection from any camera or video
Detecting people from thousands of cameras in real time
Vehicle Recognition
  1. Recognition of cars and extraction of attributes (body type, brand, model, color, license plate).
  2. Scanning license plates.
  3. Tracking the movement of vehicles
System Highlights
Each camera can be set individually.
99.9% — a real-time face recognition accuracy.
It takes less than 1 second to verify a human face with the criminal database*.
The system capacity can simultaneously analyze a large number of video streams.
How It Works
Several video stream processing modes:
  • real-time,
  • in the background,
  • offline.
01
The system instantly alerts the security service if the camera is intentionally closed by intruders.
02
Identification of intruders in real time with high accuracy.
03
The system remembers everyone who has ever been in the camera view. Thereby, it can be determined how many times and at what time the same person was in the video camera view.
04
05
By means of individual setting of access rights to the system, each user can set up personalized alarm notifications (for example, if one of the law violators got into the camera view).
Contacts:
Catherine Loginova
True Detection project manager
Made on
Tilda