Iris image data processing for template iris pattern generation

Patent metadata

Field	Value
Publication number	US-9396394-B1
Application number	US-201414496899-A
Country	US
Kind code	B1
Filing date	Sep 25, 2014
Priority date	Sep 25, 2014
Publication date	Jul 19, 2016
Grant date	Jul 19, 2016

Abstract

Systems, devices, methods, computer-readable media, techniques, and methodologies are disclosed for generating a template iris pattern using multiple image frames containing image data corresponding to detected light at different wavelengths along the electromagnetic (EM) spectrum including light in the infrared, near-infrared, and/or visible light bands.

First claim

That which is claimed is:

1. A method, comprising:
receiving, at a computer processor, first image data and second image data, wherein the first image data corresponds to an infrared or near-infrared light image and the second image data corresponds to a visible light image;
generating, by the computer processor, a first Laplacian pyramid using the first image data, wherein generating the first Laplacian pyramid comprises generating third image data comprising a first light intensity value for a first pixel and generating fourth image data comprising a second light intensity value for a second pixel;
generating, by the computer processor, a second Laplacian pyramid using the second image data, wherein generating the second Laplacian pyramid comprises generating fifth image data comprising a third light intensity value for a third pixel corresponding to the first pixel and generating sixth image data comprising a fourth light intensity value for a fourth pixel corresponding to the second pixel, wherein the third image data and the fifth image data are respectively associated with first corresponding levels of the first Laplacian pyramid and the second Laplacian pyramid and the fourth image data and the sixth image data are respectively associated with second corresponding levels of the first Laplacian pyramid and the second Laplacian pyramid;
generating, by the computer processor, a composite Laplacian pyramid, wherein generating the composite Laplacian pyramid comprises:
generating first composite image data using the third image data and the fifth image data at least in part by selecting a higher value of the first light intensity value or the third light intensity value to obtain a first selected value and associating the first selected value with a fifth pixel corresponding to the first pixel and the third pixel, and
generating second composite image data using the fourth image data and the sixth image data at least in part by selecting a higher value of the second light intensity value or the fourth light intensity value to obtain a second selected value and associating the second selected value with a sixth pixel corresponding to the second pixel and the fourth pixel;

generating, by the computer processor, output image data using the first composite image data and the second composite image data at least in part by up-sampling the first composite image data to obtain up-sampled image data and combining the up-sampled image data with the second composite image data;
identifying, by the computer processor, an iris region in the output image; and
performing, by the computer processor, feature extraction processing on the iris region to generate a template iris pattern for use in biometric authentication.

2. The method of claim 1 , wherein generating the first Laplacian pyramid further comprises:
applying, by the computer processor, a Gaussian filter to the first image data to obtain first low-pass filtered image data;
band-pass filtering, by the computer processor, the first image data by subtracting the first low-pass filtered image data from the first image data to obtain the third image data;
down-sampling, by the computer processor, the first low-pass filtered image data to obtain source image data having a smaller number of pixels than the first image data;
applying, by the computer processor, the Gaussian filter to the source image data to obtain second low-pass filtered image data; and
band-pass filtering, by the computer processor, the source image data by subtracting the second low-pass filtered image data from the source image data to obtain the fourth image data.

3. The method of claim 1 , wherein generating the third image data comprises low-pass filtering the first image data to obtain first low-pass filtered image data and generating the fifth image data comprises low-pass filtering the second image data to obtain second low-pass filtered image data.

4. The method of claim 3 , wherein generating the output image data further comprises:
combining, by the computer processor, the first low-pass filtered image data or the second low-pass filtered image data with the combination of the up-sampled image data and the second composite image data to obtain the output image data.

5. A method, comprising:
receiving, at a computer processor, first image data and second image data, wherein the first image data corresponds to an infrared or near-infrared light image and the second image data corresponds to a visible light image;
filtering, by the computer processor, the first image to obtain third image data comprising high frequency components of the first image data, wherein the third image data comprises a first light intensity value for a first pixel;
filtering, by the computer processor, the second image data to obtain a fourth image data comprising high frequency components of the second image data, wherein the fourth image data comprises a second light intensity value for a second pixel corresponding to the first pixel;
generating, by the computer processor, composite image data using the third image data and the fourth image data at least in part by selecting a higher value of the first light intensity value or the second light intensity value and associating the higher value with a third pixel corresponding to the first pixel and the second pixel; and
generating, by the computer processor, output image data using the composite image data.

6. The method of claim 5 , wherein filtering the first image data to generate the third image data comprises:
applying, by the computer processor, a low-pass filter to the first image data to obtain first low-pass filtered image data; and
applying, by the computer processor, a band-pass filter to the first image data using the first low-pass filtered image data to obtain the third image data, and
wherein filtering the second image data to obtain the fourth image data comprises:
applying, by the computer processor, the low-pass filter to the second image data to obtain second low-pass filtered image data; and
applying, by the computer processor, the band-pass filter to the second image data using the second low-pass filtered image data to obtain the fourth image data.

7. The method of claim 6 , further comprising:
down-sampling, by the computer processor, the first low-pass filtered image data to obtain source image data;
applying, by the computer processor, the low-pass filter to the source image data to obtain third low-pass filtered image data; and
applying, by the computer processor, the band-pass filter to the source image data using the third low-pass filtered image data to obtain fifth image data,
wherein the third image data and the fifth image data correspond to first and second levels, respectively, of a first Laplacian pyramid associated with the first image data.

8. The method of claim 7 , wherein the source image data is first source image data, the method further comprising:
down-sampling, by the computer processor, the second low-pass filtered image data to obtain second source image data;
applying, by the computer processor, the low-pass filter to the second source image data to obtain fourth low-pass filtered image data; and
applying, by the computer processor, the band-pass filter to the second source image data using the fourth low-pass filtered image data to obtain sixth image data,
wherein the fourth image data and the sixth image data correspond to first and second levels, respectively, of a second Laplacian pyramid associated with the second image data.

9. The method of claim 8 , wherein the third image data and the fifth image data are stored at a first location and a second location, respectively, in a first ordered data structur

Assignees

• Amazon Tech Inc

Inventors

• Solh Mashhour
• Hatipoglu Serkan

Classifications

• H04N23/11

  for generating image signals from visible and infrared light wavelengths · CPC title

• G06V40/197 Primary

  Matching; Classification · CPC title

• G06V10/52

  Scale-space analysis, e.g. wavelet analysis (multi-scale boundary representations G06V10/42) · CPC title

• G06V10/143

  Sensing or illuminating at different wavelengths · CPC title

• G06K9/00617 Primary

  Physics · mapped topic