| 摘要 |
PURPOSE: A blood vessel recognition system and method is provided to compensate for a moving or a rotation of a hand by using a hand position, continuously input hand images and a gray scale distribution when an extracted blood pattern of a back of a hand is extracted so that it can reduce an error in a person identification. CONSTITUTION: The method comprises several steps. A user fixes a hand at a main knob and an assistant knob, makes a main camera take an image of a back of the hand, and makes an assistant camera take an image of fingers(S1). A microprocessor senses a position of the hand, continuously analyzes the input hand images, and senses a moving or a rotation of the hand by using a gray scale distribution(S2). The microprocessor compensates for the moving or the rotation of the hand on a basis of the reference data stored at a database(S3), and extracts a blood vessel pattern and its feature data from the static image data transmitted via a frame grabber and a memory(S4). The microprocessor compares the extracted blood vessel pattern and its feature data with those stored at a database(S5). In a case that the extracted blood vessel pattern and its feature data are the same as those stored at a database, the microprocessor permits the user to pass an entrance or release a security system(S6). Otherwise, the microprocessor rejects a request of the user(S7).
|
