Friday, December 18, 2009

Face Recognition SDK


The FACEPORT.OCX ACTIVEX allows fast use, development and implementation of facial recognition capabilities in existing or newly designed applications. The component provides an easy to use Methods, Properties and Events that simplify the facial recognition development process.

The FACEPORT ActiveX was specially designed for embedded environments like Intel ATOM CPU, and runs very well on Low performance CPU as well as high performance CPUs. The component supports Web Cameras, IP or Analog cameras through external an injection mechanism. The component includes internal web camera management. The component contains an internal Database that provides fast yet reliable management of the matching subjects.

FACEPORT SDK

The FACEPORT FRS SDK allows third-party developers to implement our advanced facial biometric technology within their own security applications. The SDK provides tracking, enrollment, verification, classification, database, communication, and multimedia controls.

Applications

Security

  • Access Control
  • Suspect Detection
  • Time Attendance
  • Web Login
  • PC Login

Commercial

  • Repeated clients
  • Interactive Content
  • Customer counting
  • Graphical Applications

Licensing

  • PC based

SCREEN SHOT OF ACTIVEX RUNNING

Face Recognition Database

  • Jet 4.0 - MS Access database (Optional: ODBC)
  • Log Database : Jet 4.0 MS Access database
  • Provides all interfaces necessary to Add, Delete and Retrieve biometric templates, images and users from the central database.

Properties

Name of Property Description
NoOfFramesForValidation Min. number of frames for validation
NoOfFramesForEnrollment Min. number of frames for enrollment
FaceConfidenceThreshold 
MinimalIOD Min. Inner distance between 2 eyes (Default 60)
MaximalIOD Max. Inner distance between 2 eyes (Default 1500)
FaceQualityThreshold 
MatchingThreshold Min. value for matching
FAR Will be calculated from matching threshold (theoretical value)
MatchingAttempts number of matching attempts
AllowLiveness Boolean - will define if we need to use liveness test
LivenessThreshold how strict to check for Liveness in an image.(default 50)
MaxRecordsPerTemplate 
AllowAging Boolean - will define if we need to use aging+
AgingMinimalThreshold 
EnrolledFaceImagePath 
RejectedFaceImagePath 
ValidatedFaceImagePath 
LogDBPath 
FRSDBPath 
IsRegistered returns true if registered
FRR Will be calculated from matching threshold (theoretical value)
Camera Gets/Sets the currently selected Camera
VideoFormat Gets/Sets the Currently Selected Camera’s Video Format.
FlipImagesHorizontal Gets/Sets value whether to Flip the Images Horizontally
SaveEnrolledFaceImage Gets/Sets value Whether to Save the Learned/Enrlolled Face Image to harddisk
SaveRejectedFaceImage Gets/Sets value Whether to Save the Rejected Face Image to harddisk
SaveValidatedFaceImage Gets/Sets value Whether to Save the Validated/Matched Face Image to harddisk

Events

Name of Events Description
CameraOpened() 
CameraClosed() 
NewCameraFrame(bitmap image) for every new frame
ExceptionHappened(string Description, int ErrorNumber) 
FaceLearnt(int FaceQuality, float Similarity, float FaceConfidence, int IOD) after new enrollment
FaceValidated(int FaceQuality, float Similarity, float FaceConfidence, int IOD) after validation
FaceLearningFailed(int ErrorTypeID, String Description, int FaceQuality, float Similarity, float FaceConfidence, int IOD ) after failure of enrollment
FaceValidationFailed(int ErrorTypeID, String Description, int FaceQuality, float Similarity, float FaceConfidence, int IOD)after failure of validation
FaceDeleted() after deletion of face
LogOverflow() when log record number is above 500K records

Methods

Name of Method Description
InitializeFRS()  
OpenFRSSettingsDialog() open face recognition component configuration
Bool LearnFaceFromCamera(string FaceID) 
Bool LearnFaceFromImage(string FaceID, string ImagePath) 
Bool ValidateFace(string FaceId) 
Bool OpenCamera(string camera) 
String[] GetRejectedImagesByFaceID(string FaceID) 
String[] GetRejectedFaceIDsFromDate(string FromDate, string ToDate) 
DeleteRejecedImagesFromDate(string FaceID, string UpToDate) 
DeleteAllRejecedImagesFromDate(string UpToDate) 
String[] GetImagesByFaceID(string FaceID) 
DeleteFaceID(string FaceID) 
Bool DeleteFaceIDImages(string FaceID, string UpToDate) 
Float MatchImageWithFace(string filename, string FaceID) between the two faces
AddToLog(string timestamp, string type, string data) 
ClearLog() Clear Log db
String[] GetConnectedCameras() 
CloseCamera() 
ShowCameraPreviewDialog(int x,y)- -1,-1 center of screen
bool SaveSnapshot(string Filename) 
PlaceImageAnnotation(string TargetFilename, string Data) 
InjectImage(ByVal HImage As Long) Injects an Image using it HBITMAP or Image
HandleInjectImageFromFile(ByVal fileName As String) Injects an Image from a File

No comments:

Post a Comment