System and method for measuring clarity of images used in an iris recognition system

• US 8,260,009 B2
• Filed: 08/15/2008
• Issued: 09/04/2012
• Est. Priority Date: 08/15/2007
• Status: Active Grant

• Current Assignee: Indiana University Research & Technology Corporation (Indiana University)
• Original Assignee: Indiana University Research & Technology Corporation (Indiana University)
• Inventors: Du, Yingzi, Belcher, Craig S.
• Primary Examiner(s): Mehta, Bhavesh
• Assistant Examiner(s): Park, Edward
• Application Number:
  US12/673,626
• Publication Number:
  US 20100202668A1
• Time in Patent Office: 1,481 Days
• Field of Search: 382/115, 382/117, 382/127, 235/380, 340/5.1, 340/5.92, 704/246, 704/273, 902/3, 902/4, 902/25, 351/200, 351/204, 283/68, 356/71
• US Class Current: 382/117
• CPC Class Codes: G06V 10/993 Evaluation of the quality of the acquired pattern
• CPC Class Codes: G06V 40/193 Preprocessing; Feature extraction

Abstract

An image quality measuring method enables a biometric image to be evaluated to determine whether the biometric image data are adequate for identification processing. The method includes converting a biometric image to dimensionless image data, filtering the dimensionless image data with a band pass filter, identifying a plurality of portions in the filtered data as containing identification features, each portion in the plurality having an information measurement that indicates feature content greater than portions in the filtered data that are excluded from the plurality, and measuring clarity for the biometric image from the identified plurality of portions in the filtered data.

Claims

1. A method for measuring quality of a biometric image comprising:

converting a biometric image to dimensionless image data;

filtering the dimensionless image data with a band pass filter;

applying a sliding window to the filtered dimensionless image data to identify portions for information distance calculations;

generating an information distance for the portions of the filtered data identified by the sliding window;

weighting information distances for selected portions identified by the sliding window;

ranking the weighted information distances;

selecting a number N portions corresponding to the N weighted information distances that are larger than the other weighted information distances in the ranking of the weighted information distances to identify a plurality of portions in the filtered dimensionless image data as containing identification features; and

measuring clarity for the biometric image from the identified plurality of portions in the filtered dimensionless image data.

2. The method of claim 1, the biometric image conversion further comprising:

converting biometric image data into polar coordinate image data.

3. The method of claim 2 further comprising:

converting the dimensionless image data to spatial domain data; and

converting the spatial domain data to dimensionless data following application of a Log-Gabor band pass filter to the spatial domain data.

4. The method of claim 1, the biometric image conversion further comprising:

converting biometric image data into log-polar coordinate image data.

5. The method of claim 1, the filtering of the dimensionless image data further comprising:

applying a Log-Gabor filter to the dimensionless image data.

6. The method of claim 1 further comprising:

measuring occlusion in the dimensionless biometric image.

7. The method of claim 6 further comprising:

measuring dilation in the dimensionless biometric image.

8. The method of claim 7 further comprising:

generating a quality measurement for the biometric image from the clarity measurement, the occlusion measurement, and the dilation measurement.

9. A system for measuring quality of a biometric image comprising:

a preprocessor configured to convert biometric image data to one of polar coordinate data and log-polar coordinate data to form dimensionless image data;

a feature extractor having a domain converter configured to convert the dimensionless image data to spatial domain data, a Log-Gabor filter for filtering the spatial domain data, and an inverse domain generator configured to convert the filtered spatial domain data to the dimensionless data coordinates from which the domain converter converted the dimensionless image data for filtration;

a region of interest selector configured to identify a plurality of portions of the filtered image data, each portion in the plurality of portions having an information distance that is indicative of feature content, and to weight the information distances for portions selected from the plurality of portions to enable the feature extractor to rank the weighted information distances measured for the selected portions; and

a feature information measurement generator configured to measure clarity of the biometric image from the ranked weighted information distances measured for the selected portions in the plurality of portions.

10. The system of claim 9, the region of interest selector further comprising:

a sliding window to select the portions of the filtered dimensionless data for information distance calculations; and

a information distance generator for measuring the information distance for each portion selected by application of the sliding window to the filtered dimensionless data.

11. The system of claim 9 further comprising:

an occlusion measurement generator configured to measure occlusion in the biometric dimensionless image data; and

a dilation measurement generator configured to measure dilation in the biometric dimensionless image data.

12. The system of claim 11 further comprising:

an image quality measurement generator for generating a quality measurement for the biometric image from the clarity measurement, the occlusion measurement, and the dilation measurement.