Method and arrangement for optical recording of data
AssigneeTBS Holding AG
InventorsRudolf Hauke, Giuseppe Parziale, Gerhard Paar, Peter-Michael Merbach
Life of this patent
Estimating the term…
Abstract
The invention relates to a method and a system for recording biometric data, especially for recording the features of the fingers or faces of persons, wherein an object is recorded by optically scanning it and evaluated by digital image processing. The inventive method is characterized in that the object is recorded simultaneously or almost simultaneously by at least two different recording devices, whereby all pixels of the surface to be imaged are imaged in at least two different directions in at least one digital two-dimensional image each and a three-dimensional model of the viewed object is calculated from at least two images.
Claims11
A method for acquiring biometric data of an object comprising:
- acquiring a first digital two-dimensional image of a first set of points on a surface of the object from a first taking direction using a first optical scanning sensor;
- acquiring a second digital two-dimensional image of the first set of points from a second taking direction using a second optical scanner at least approximately simultaneously with the acquiring of the first image;
- calculating a three-dimensional model of the object using the first and second digital two-dimensional images, wherein the calculating is performed using a geometric basic model of the object, the geometric basic model being simplified in comparison to the object, and wherein height deviations between the geometric basic model and the object are determined, the height deviations being orthogonal to a local surface of the geometric basic model;
- acquiring a first calibration image of a set of defined points on a surface of a calibration object from the first taking direction using the first optical scanning sensor;
- and
- acquiring a second calibration image of the set of defined points from the second taking direction using the second optical scanning sensor at least approximately simultaneously with the acquiring of the first image.
A method for acquiring biometric data of an object comprising:
- acquiring a first digital two-dimensional image of a first set of points on a surface of the object from a first taking direction using a first optical scanning sensor;
- acquiring a second digital two-dimensional image of the first set of points from a second taking direction using a second optical scanner at least approximately simultaneously with the acquiring of the first image;
- calculating a three-dimensional model of the object using the first and second digital two-dimensional images;
- acquiring a third digital two-dimensional image of a second set of points on a surface of the object from a third taking direction using a third optical scanning sensor, wherein the three-dimensional model includes a model of one of the surface and the image texture of the surface;
- acquiring a fourth digital two-dimensional image of the second set of points from a fourth taking direction using a fourth optical scanner at least approximately simultaneously with the acquiring of the third image, and wherein the calculating of the three-dimensional model of the object is performed using the first, second, third, and fourth digital two-dimensional images;
- recording the two-dimensional images and determining two-dimensional parallaxes for the stereoscopic analysis from the recorded two-dimensional images;
- and
- pre-undistorting the two-dimensional images using a geometric basic model before determining the parallaxes, the geometric base model being simplified in comparison to the object, wherein the object includes a finger, and wherein the basic model includes at least a portion of a solid of revolution, radii of the solid of revolution being approximated to the shape of the finger depending on a position of the radii along an axis of the solid of revolution.
A method for acquiring biometric data of an object comprising:
- acquiring a first digital two-dimensional image of a first set of points on a surface of the object from a first taking direction using a first optical scanning sensor;
- acquiring a second digital two-dimensional image of the first set of points from a second taking direction using a second optical scanner at least approximately simultaneously with the acquiring of the first image;
- calculating a three-dimensional model of the object using the first and second digital two-dimensional images;
- acquiring a first calibration image of a set of defined points on a surface of a calibration object from the first taking direction using the first optical scanning sensor;
- acquiring a second calibration image of the set of defined points from the second taking direction using the second optical scanning sensor at least approximately simultaneously with the acquiring of the first image;
- determining an inner and an outer orientation of each of the first and second optical scanning sensors from the first and second calibration images;
- calculationally converting the three-dimensional model to a plane area, wherein the calculationally converting the three-dimensional model includes simulating an unrolling movement of a three-dimensional representation and an orthogonal projection of the three-dimensional model onto a profile body and assigning points of the profile body's surface to points of the plane area, and wherein the profile body includes a half cylinder, and wherein a length of an arc of the half cylinder is assigned to a linear coordinate in the plane area.
Dependencies are inferred from the claim text.