MONROE COUNTY SHERIFF’S OFFICE

General Order

1. PURPOSE: The purpose of this General Order is to establish guidelines for using biometric identification technology, specifically facial recognition.

   
   

2. DISCUSSION: In 2002 the Pinellas County Sheriff’s Office (PCSO) implemented facial recognition as a biometric tool used to identify or verify individuals as part of the law enforcement mission. PCSO established and hosts one of the largest collaborative facial recognition systems in Florida and allows access to authorized law enforcement agencies. Facial recognition search capabilities are shared with other law enforcement users as well as partner agencies who actively contribute facial images to the database for law enforcement purposes only. The Face, Analysis, Comparison, and Examination System (FACESNXT) is hosted through FDLE’s CJNET, and access is only granted to authorized agencies. Authorized agencies are those that have submitted formal requests and executed a user agreement memorandum of understanding (MOU). Monroe County Sheriff’s Office (MCSO) has executed an MOU, and FACESNXT can be accessed by authorized users via the link located on MCSO intranet page (black page) or CJNet.

   
   

3. DEFINITIONS:

   
   

   1. Automated Face Recognition: The automated search of a facial image in a biometric database (one-to-many), which typically results in a group of facial images ranked by computer-evaluated similarity.

      
      

   2. Biometrics: Distinctive and measurable human characteristics that can be used to identify people apart from demographic data like name and date of birth. Fingerprints and facial features are examples of commonly used biometrics. Since biometrics are unique to individuals, they are more reliable in verifying identity than knowledge based methods

      
      

   3. Enroll: The act of capturing a facial image, creating a template and entering the template into a facial recognition gallery.

      
      

   4. Facial Identification: A task where the face recognition system searches a database for a reference matching a submitted face image (probe) and, if found, returns a possible corresponding identity.

      
      

   5. Facial Verification: Manual verification is the process by which a user conducts a one-to-one comparison between a probe and gallery face pair. Automated verification is the process by which the facial recognition system compares a known and unknown face pair to generate a resulting similarity score.

      
      

   6. Gallery: The result of an automated facial recognition search. The gallery is a rank ordered grid of potential matching candidates to the face image (probe) submitted for search.

      
      

   7. Probe: The facial image or template searched against the gallery in an automated facial recognition system.

   8. Template: Mathematical representation of facial features prepared from a face image.

      
      

4. POLICY:

   
   

   1. MCSO may utilize facial recognition for law enforcement investigations requiring identification or verification of subjects where a digital image is available. Facial recognition searches and comparisons may be performed through the user’s web browser by accessing the Face Analysis, Comparison, and Examination System (FACESNXT). The FACESNXT application is available through MCSO intranet page (black page) or CJNet and is to be used to assist in law enforcement investigations. Information found through a facial recognition search is to be considered a lead for further investigation.

      
      

   2. FACESNXT will only be used for official criminal justice purposes.

      
      

   3. All users are required to complete the online training modules prior to gaining access to the FACESNXT application.

      
      

   4. Users may submit digital face images (probes) in JPG, BMP, and PNG file formats to FACESNXT.

      
      

   5. Users are required to include a reason for search. This reason shall be the associated MCSO/HIDTA case number for the subject of the facial recognition search.

      
      

   6. Users will review the FACESNXT rank ordered gallery for potential matches to the submitted probe.

      
      

   7. Users may retrieve demographic information from subjects selected within the FACESNXT gallery.

      
      

   8. Users may perform detailed side-by-side facial image comparisons on selected probe and gallery images to assist in an investigative lead determination.

      
      

   9. The user will make the determination of an investigative lead of the probe image to the gallery image(s).

      
      

   10. Any information found through facial recognition search is for lead purposes only.

       
       

   11. FACESNXT is an investigative tool, and any law enforcement action taken based on a submission to FACESNXT shall be based on the officer’s own identity determination and not solely the results of a FACESNXT search.

       
       

   12. The agency administrator will conduct a quarterly FACESNXT user audit to ensure adherence to this policy. All user activity is logged.

       
       

   13. The agency administrator, who is the Major of Operations, or his designees will update user access immediately if a user is terminated, retires, or no longer needs FACESNXT.

       
       

   14. Users will bring any problems with said equipment or FACESNXT application to the immediate attention of the agency administrator.