Software Application for Assisted Categorization, Processing, and Formatting of Biometrics
Forensic Workbench is a software application that utilizes several Aware SDKs for the categorization, processing, and standards-compliant formatting of biometric images and demographic data. The primary function of the application is the easy assembly of disparate biometric imagery and text into a standard-compliant data structure, such as ANSI/NIST-ITL 1-20xx, FBI EFTS, and DoD EBTS.
Forensic Workbench provides the operator the ability to organize a collection of biometric and demographic images and text files. The basic capabilities include:
- Categorization of data into face, finger, iris, and documents,
- Processing of finger face and iris imagery to specifically label each image and sub-image as facial frontal or other profile, and fingerprint as to its source finger,
- Measurement and reporting of image quality metrics for each categorized and labeled sub-image,
- Generation of an ANSI/NIST-ITL-compliant output file (eg. DOD EBTS 1.2, Interpol 4.x, FBI EFTS), and
- Modification of the data within an existing EBTS file.
The first operation in the application workflow is to identify a project directory; all supported files within this directory are loaded. The operator can then categorize each image (i.e. label it as a finger, face, iris, or document). Once the categorization is complete, image type-specific operations may take place.
Step 1 – Data Selection. The left side shows the raw biometric data yet to be processed. The right side shows biometric data the operator has selected for processing.
Step 2 - Assembly. This stage allows the operator both automated and manual options to analyze and process the biometric images so that they are of acceptable format and quality.
Forensic Workbench can display and process facial and fingerprint imagery. Facial images can be cropped to identify facial bounding boxes. The crop regions can then be labeled as either frontal, left profile, left 45 profile, right profile, and right 45 profile. The operator may automatically extract frontal images from an input image or manually identify a face via the marking of the eye centers. For frontal images, Aware PreFace is used to identify facial features and ensure compliance to an input frontal face specification. Input fingerprint images can be cropped to identify each finger (right and left thumb, index, middle, ring, little). Each finger can be labeled with its source and impression type (live scan, ink, rolled, plain) and its quality reported (Aware QualityCheck and NFIQ). The image storage format can be identified (RAW or WSQ).
Demographic data can also be entered. The demographic fields are fully configurable in accordance to the output ANSI/NIST specification. The entry of demographic data is supported through the display of image files and text files for operator review.
Once data categorization and processing has been completed, the operator may generate an ANSI/NIST transaction. All data is validated via NISTPack prior to generation of the file.