|
Jupiter DVR Applications |
|||||||||||||||||||||||||||
Our next generation digital video recording systems are equipped with a family of advanced image processing applications to give you no-compromise quality, reliability and functionality you come to expect and demand from Jupiter technology. Power your infrastructure with advanced OCR and FRS solutions from Jupiter. They are guaranteed to delight your customers, improve security and reduce your operating cost. |
||||||||||||||||||||||||||||
|
FR (Face Recognition) Solution |
||||||||||||||||||||||||||||
Face recognition is popular because of its non-intrusive nature (it does not require the cooperation of the subject) and are applied in many exciting and mission critical applications such as law enforcement, premise access control for casinos, airlines, airports, immigration and customs. Jupiter FRS (face recognition) solution provides an integrated functionality of digital video recording and face recognition using facial biometrics encoding to produce face encode array. The encode arrays are used to query facial encode array database located locally or over a network for subject identification (who am I?) or verification (I am who?). The FR Solutions are:
|
 |
|||||||||||||||||||||||||||
|
Jupiter FRS Encoding Process |
||||||||||||||||||||||||||||
Jupiter FRS (face recognition) solutions employ the world-leading ID-2000® technology. The ID-2000 technology uses a 3-D deformable surface model and utilizes a two stage encoding process, first to locate the face and features and second to create a unique identifier for the individual’s face. The encoding process
applies deformations within the spartial orientations of pitch, roll and
yaw of the head. The surface is re-rendered to different light source
positions to achieve the best fit to the face. Once best fit is achieved
for the whole face, selected sub-regions of the face areas that are least
affected by environmental changes (eg. Beard growth, hair style change
or weight changes) are used to perform a more localized fine tune fit.
The face is then normalized using a complex location and normalization
factors consisting of positions of the tip of the nose, left and right
eyes called anchor points. Finally a spectral analysis is performed using
these anchor points to generate a descriptor array of 117 identifiers
that are unique as a whole. The descriptor array is combined with other
information to form the encode array. |
||||||||||||||||||||||||||||
|
Technical Specification |
||||||||||||||||||||||||||||
ID-2000
Version 9.0 test results using 16,000 images in DB
O/S:
Supports Windows 98, Windows NT, Windows 2000 & XP Download PDF brochures! |
||||||||||||||||||||||||||||
