G06V 40/40
Definition
Diese Klassifikationsstelle umfasst:(Für diese Definition ist die deutsche Übersetzung noch nicht abgeschlossen)
Spoof detection, i.e. detecting an attempt to fool a biometric system by presenting data which is not genuine. An example is the detection of inanimate replicas of living tissue, and the distinguishing of such replicas, e.g. a rubber model of a finger, from parts of living beings.
Spoof detection can be performed using acquisition arrangements in which the sensor is provided with specialised hardware to assess or highlight the genuineness of the acquired data (e.g. using special illumination in infrared) or by performing image processing operations (e.g. colour analysis to discriminate the genuine skin against a copy). Multiple biometric modalities can be involved:
- Signals such as blood pressure, pulse and perspiration at the fingertips, hippus movement of the pupil, brain waves (EEG) and electrical heart signals (ECG) in combination with other biometric images;
- Reflexive signals such as pupillary light reflex (pupil dilation), corneal reflex (blink reflex) and patellar reflex (knee-jerk);
- Voluntary signals given unconsciously or as a response to a "challenge” such as blinking, mouth movements and facial expressions.
Other properties of a body can be assessed:
- Determination of the flatness of a face to detect use of a picture to challenge a biometric system (“spoof-by-picture”);
- Light distribution in a real finger which differs from a fake finger.
Examples
Recognition method using hand biometrics with anti-counterfeiting. The user is asked to perform randomly selected gestures with the hand, e.g. rotate the hand to the left, clench into a fist. The gestures are recognised, allowing the method to determine that a real user is standing in from of the camera.
When the eye opens, the eye aspect ratio is roughly constant, and only fluctuates around the range 0.25. Once the eye blinks and closes, because the vertical distance is almost zero, the eye aspect ratio is correspondingly reduced to zero. When the eye opens again, the eye aspect ratio rises to the range 0.25 again. These measurements may indicate if the person is a real, genuine person or a fake.
Beziehungen zu anderen Klassifikationsstellen
Authentication of users for computer access is classified in group G06F 21/32.
Authentication of financial documents is classified in group G07D 7/00.
Querverweise
Informative Querverweise
| Detection or correction or errors, e.g. by rescanning the pattern; Evaluation of the quantity of an acquired biometric pattern | G06V 10/98 |
| Recognition of fingerprints or palmprints in images or video | G06V 40/12 |
| Recognition of vascular patterns in images or video | G06V 40/14 |
| Recognition of human faces in images or video, e.g. facial parts, sketches or expressions | G06V 40/16 |
| Recognition of eye characteristics in images or video, e.g. of the iris | G06V 40/18 |
| Recognition of movements or behaviour in images or video, e.g. gesture recognition | G06V 40/20 |
| Recognition of signatures | G06V 40/30 |
| Multimodal biometrics, e.g. combining information from different biometric modalities | G06V 40/70 |
| Fittings or systems for preventing or indicating unauthorised use or theft of vehicles | B60R 25/00 |
| Digitisers as the input arrangement for user-computer interaction, e.g. touch screens or touch pads | G06F 3/041 |
| User authentication using biometric data for protecting computers, components thereof, programs or data against unauthorised activity | G06F 21/32 |
| User authentication by graphic or iconic representation for protecting computers, components thereof, programs or data against unauthorised activity | G06F 21/36 |
| Testing specially adapted to determine the identity or genuineness of valuable papers | G07D 7/00 |
| Checking-devices for individual entry or exit registers | G07C 9/00 |
| Arrangements for secret or secure communications; Network security protocols | H04L 9/00 |
| Means for preventing unauthorised calls from a telephone set | H04M 1/667 |
Spezielle Klassifizierungsregeln
This group is used alone when no technical contribution can be identified in the processing associated with biometric authentication. If, however, a technical contribution can be identified in biometric authentication, the respective groups are allocated in combination with this group. In other words, anti-spoofing is usually a part of an authentication process which acts as a verifier of liveliness, thus anti-spoofing inventions often rely on processing biometric data of a certain modality provided in the following groups:
- Fingerprints or palmprints – group G06V 40/12;
- Human faces, e.g. facial parts, sketches or expressions – group G06V 40/16;
- Eyes – group G06V 40/18;
- Movements or behaviour, e.g. gesture recognition – group G06V 40/20;
- Signatures – group G06V 40/30.
For example, in order to assure safe biometric authentication, the face matching process classified in group G06V 40/16 combined with a liveness detection, e.g. by determining if the user in front of the camera is moving his mouth when requested so that a spoof-by-picture attack can be prevented, is classified also in group G06V 40/40.
G06V 40/40
Definition Statement
This place covers:Spoof detection, i.e. detecting an attempt to fool a biometric system by presenting data which is not genuine. An example is the detection of inanimate replicas of living tissue, and the distinguishing of such replicas, e.g. a rubber model of a finger, from parts of living beings.
Spoof detection can be performed using acquisition arrangements in which the sensor is provided with specialised hardware to assess or highlight the genuineness of the acquired data (e.g. using special illumination in infrared) or by performing image processing operations (e.g. colour analysis to discriminate the genuine skin against a copy). Multiple biometric modalities can be involved:
- Signals such as blood pressure, pulse and perspiration at the fingertips, hippus movement of the pupil, brain waves (EEG) and electrical heart signals (ECG) in combination with other biometric images;
- Reflexive signals such as pupillary light reflex (pupil dilation), corneal reflex (blink reflex) and patellar reflex (knee-jerk);
- Voluntary signals given unconsciously or as a response to a "challenge” such as blinking, mouth movements and facial expressions.
Other properties of a body can be assessed:
- Determination of the flatness of a face to detect use of a picture to challenge a biometric system (“spoof-by-picture”);
- Light distribution in a real finger which differs from a fake finger.
Examples
Recognition method using hand biometrics with anti-counterfeiting. The user is asked to perform randomly selected gestures with the hand, e.g. rotate the hand to the left, clench into a fist. The gestures are recognised, allowing the method to determine that a real user is standing in from of the camera.
When the eye opens, the eye aspect ratio is roughly constant, and only fluctuates around the range 0.25. Once the eye blinks and closes, because the vertical distance is almost zero, the eye aspect ratio is correspondingly reduced to zero. When the eye opens again, the eye aspect ratio rises to the range 0.25 again. These measurements may indicate if the person is a real, genuine person or a fake.
Relationships with other classification places
Authentication of users for computer access is classified in group G06F 21/32.
Authentication of financial documents is classified in group G07D 7/00.
References
Informative references
| Detection or correction or errors, e.g. by rescanning the pattern; Evaluation of the quantity of an acquired biometric pattern | G06V 10/98 |
| Recognition of fingerprints or palmprints in images or video | G06V 40/12 |
| Recognition of vascular patterns in images or video | G06V 40/14 |
| Recognition of human faces in images or video, e.g. facial parts, sketches or expressions | G06V 40/16 |
| Recognition of eye characteristics in images or video, e.g. of the iris | G06V 40/18 |
| Recognition of movements or behaviour in images or video, e.g. gesture recognition | G06V 40/20 |
| Recognition of signatures | G06V 40/30 |
| Multimodal biometrics, e.g. combining information from different biometric modalities | G06V 40/70 |
| Fittings or systems for preventing or indicating unauthorised use or theft of vehicles | B60R 25/00 |
| Digitisers as the input arrangement for user-computer interaction, e.g. touch screens or touch pads | G06F 3/041 |
| User authentication using biometric data for protecting computers, components thereof, programs or data against unauthorised activity | G06F 21/32 |
| User authentication by graphic or iconic representation for protecting computers, components thereof, programs or data against unauthorised activity | G06F 21/36 |
| Testing specially adapted to determine the identity or genuineness of valuable papers | G07D 7/00 |
| Checking-devices for individual entry or exit registers | G07C 9/00 |
| Arrangements for secret or secure communications; Network security protocols | H04L 9/00 |
| Means for preventing unauthorised calls from a telephone set | H04M 1/667 |
Special rules of classification
This group is used alone when no technical contribution can be identified in the processing associated with biometric authentication. If, however, a technical contribution can be identified in biometric authentication, the respective groups are allocated in combination with this group. In other words, anti-spoofing is usually a part of an authentication process which acts as a verifier of liveliness, thus anti-spoofing inventions often rely on processing biometric data of a certain modality provided in the following groups:
- Fingerprints or palmprints – group G06V 40/12;
- Human faces, e.g. facial parts, sketches or expressions – group G06V 40/16;
- Eyes – group G06V 40/18;
- Movements or behaviour, e.g. gesture recognition – group G06V 40/20;
- Signatures – group G06V 40/30.
For example, in order to assure safe biometric authentication, the face matching process classified in group G06V 40/16 combined with a liveness detection, e.g. by determining if the user in front of the camera is moving his mouth when requested so that a spoof-by-picture attack can be prevented, is classified also in group G06V 40/40.