| 摘要 |
A face region detecting apparatus and method comprising a face detection and recognition unit that (a) extracts a face image from an inputted image, (b) encodes the face image into a "feature code" and (c) stores the feature code in a database; and another unit/piece of apparatus (a face detection server in claim 1 or a face recognition module in claim 9) which (d) compares the face feature code with previously stored face feature codes, (e) determines whether the features of the codes match and (f) transmits a recognition result to the face detection and recognition unit. The inputted image may be from a digital video recorder, particularly one which records in infrared (IR) and has an infrared iris filter. The apparatus may comprise an light emitting diode (LED) emitting infrared light to illuminate subject people. The extraction of the face image may comprise searching and measuring information on a front face in real time, for example by: extracting a face screen from the video data input; deriving data values from the face screen by repeating course and fine inspections using an 'Ada Boost' calculation method; using these to determine face features; using these to determine geometrical positions of parts of the face and to derive a final face screen by equalising variation of brightness between left and right sides of the face; and deriving an image matching to a standard face screen by rotating, enlarging, and reducing the final face screen. |
