Method and system for increasing biometric acceptance rates and reducing false accept rates and false rates
View Patent ↗A method and system for increasing biometric acceptance rates is provided. The invention creates numerous enrollment reference images so that when an individual presents them self to a biometric sensor, the individual's live image is compared to numerous enrollment reference images which establish a very high matching threshold score and at least one reference image which exceeds the matching threshold for an individual. Since the acceptance threshold is set very high, around 0.98 the chances that a false accept occurs is significantly reduced. By selecting numerous different reference photographs of the enrolled candidate, there's a greater possibility of a very high score match to the live scan. As you increase the number of biometric enrollment reference photos, you increase the probability of the live scan achieving a very high match against the enrollment reference photos. The enrollment reference photos may be visualized as a shotgun distribution on a graph.
1. A method for reducing both false acceptance rates and false rejection rates in a biometric matching process, the method comprising:
a. enrolling a plurality of people into an enrollment reference database of photographic images by filming a body part of each person of the plurality of people with a video camera while the body part or video camera is moving, wherein large numbers of images of the person's body part at multiple, varying angles are captured in video;
b. extracting entire individual video frames and placing the extracted individual video frames, without segmentation into regions, into a shotgun file for each said person of the plurality of people through a specified time period or percent of face rotations and storing the shotgun file including the individual video frames into the enrollment reference photographic image database for each said person of the plurality of people
c. using software which converts said extracted individual video frames of each said shotgun file into body part matching templates;
d. collecting at least one live biometric body part image and converting said at least one biometric live body part image to a biometric matching template
e. using a matching algorithm to compare said biometric matching template with the body part matching templates to generate a match score;
f. selecting the highest match score above a very high threshold and presenting the extracted individual video frame of the highest match score that is above the very high threshold.
2. The method of claim 1 , wherein the body part is selected from the group consisting of a face, fingerprint, iris, hand, nose, and ear.
3. The method of claim 1 , wherein the at least one biometric body part image is several biometric body part images.