SYSTEM FOR OPERATIONAL VISUAL ANALYSIS FOR LAW INFORCEMENT AGENCIES
DESCRIPTION
SOVA is an automated information storage and retrieval system for facial recognition.
SOVA is designed to create and store a database of those wanted by law enforcement agencies and for the identification of citizens using biographical data, a verbal description and/or a facial image.
The facial image is entered as a search parameter. The result from the system is the automatic extraction of an image list ranked according to the degree of similarity with the input image from the database. The comparison is made using the "one-to-many" principle.
In Kazakhstan and Russia SOVA is used by investigation departments of the Ministry of Internal Affairs for the detention of felons and suspect identification by victims, witnesses or using the result of operational filming, etc.
TECHNICAL CHARACTERISTICS
Table 1.
Probability of identification and Time of search.
| № |
Database size |
Probability
identification in % |
Time of search in seconds |
| 1 |
10,000 images |
99.9 |
0.16 |
| 2 |
100,000 images |
99.2 |
0.28 |
| 3 |
1,000,000 images |
96.3 |
1.1 |
| 4 |
10,000,000 images |
87.4 |
7.08 |
Table 2.
Time of search using data entry and verbal description
| № |
Parameters |
Values |
| 1 |
Database size |
10,000,000 cards |
| 2 |
Search time by data entry(i.e. name, DOB, etc.) |
< 3 seconds |
| 3 |
Search time using verbal description |
< 3 seconds |
Performance specification
| Image encodes time: |
0.26 seconds |
| Retrieval rate: |
1.66 million images per second |
| Percent of automatic encoding: |
99.95% |
| Speed of automatic input: |
2 millions images per twenty-four-hours |
The technical characteristics were based on a Pentium IV / 2.8 GHz / 2 Gb RAM processor using MS Windows 2000/XP.
Demonstration of operation capability[Review 8,25Mb] [Download 8,61Mb]
