Method for automatically detecting digital images that are undesirable for placing in albums

• US 6,535,636 B1
• Filed: 03/23/1999
• Issued: 03/18/2003
• Est. Priority Date: 03/23/1999
• Status: Expired due to Term

• Current Assignee: Monument Peak Ventures, LLC (Dominion Harbor Enterprises, LLC)
• Original Assignee: Eastman Kodak Company
• Inventors: Loui, Alexander C., Savakis, Andreas E.
• Primary Examiner(s): Wu, Jingge
• Application Number:
  US09/274,645
• Time in Patent Office: 1,456 Days
• Field of Search: 382/162-176, 358/450-487, 359/666-670
• US Class Current: 382/173
• CPC Class Codes: G06T 7/0002 Inspection of images, e.g. flaw detection
• CPC Class Codes: G06V 10/993 Evaluation of the quality of the acquired pattern

Abstract

A method for automatically classifying a digital image as a dud, the method comprises the steps of receiving the digital image; determining individually or any combination of sharpness,;contrast, noise, and exposure of the digital image; determining a threshold individually for sharpness, contrast, noise, and exposure, or a threshold for any combination of sharpness, contrast, noise, and exposure which determined threshold or thresholds determines if the image is classified as a dud; and classifying the image as a dud based on the determination of the previous step.

Claims

1. A method for automatically classifying a digital image as a dud, the method comprising the steps of:

(a) receiving the digital image;

(b) determining individually the sharpness, contrast, noise, and exposure of the digital image;

(c) determining a threshold individually for sharpness, contrast, noise, and exposure, which determined threshold or thresholds determines if the image is classified as a dud;

(d) determining an overall image quality parameter for the digital image by forming a linear combination of the sharpness, contrast, noise and exposure thresholds; and

(e) classifying the image as a dud based on the determination of the overall image quality parameter in step (d).

2. The method as in claim 1, wherein the determining the sharpness includes the steps of:

obtaining an edge profile;

computing a histogram from the edge profile;

locating predetermined edges of the edge histogram; and

computing a centroid of gradients of the predetermined edges.

3. The method as in claim 1, wherein the step of determining the contrast includes the steps of:

obtaining an edge profile;

computing an intensity histogram from the edge profile; and

calculating a standard deviation of the histogram.

4. The method as in claim 1, wherein the step of determining the noise includes the steps of:

obtaining an edge profile;

computing a histogram from the edge profile; and

locating flat regions of the histogram; and

calculating a standard deviation of the flat regions.

5. The method as in claim 1, wherein the step of determining the exposure includes the steps of:

comparing each image pixel to either an overexposure or underexposure threshold;

computing a percentage of pixels that are either above the overexposure threshold or below the underexposure threshold; and

determining the overexposure or underexposure based on the computed percentage.

6. A system for automatically classifying a digital image as a dud, the system comprising:

(a) means for receiving the digital image;

(b) means for determining individually the sharpness, contrast, noise, and exposure of the digital image;

(c) means for determining a threshold individually for sharpness, contrast, noise, and exposure, which determined threshold or thresholds determines if the image is classified as a dud;

(d) means for generating an overall image quality parameter for the digital image by forming a linear combination of the sharpness, contrast, noise and exposure thresholds; and

(e) means for classifying the image as a dud based on the overall image quality parameter.

7. The system as in claim 6, wherein the determining the sharpness means further comprises:

means for obtaining an edge profile;

means for computing a histogram from the edge profile;

means for locating predetermined edges of the edge histogram; and

means for computing a centroid of gradients of the predetermined edges.

8. The system as in claim 6, wherein the determining the contrast means further comprises:

means for obtaining an edge profile;

means for computing an intensity histogram from the edge profile; and

means for calculating a standard deviation of the histogram.

9. The system as in claim 6, wherein the determining the noise means further comprises:

means for obtaining an edge profile;

means for computing a histogram from the edge profile; and

means for locating flat regions of the histogram; and

calculating a standard deviation of the flat regions.

10. The system as in claim 6, wherein the determining the exposure means further comprises:

means for comparing each image pixel to either an overexposure or underexposure threshold;

means for computing a percentage of pixels that are either above the overexposure threshold or below the underexposure threshold; and

means for determining the overexposure or underexposure based on the computed percentage.

11. A computer program product for automatically classifying a digital image as a dud, comprising:

computer readable storage medium having a computer program stored thereon for performing the steps of;

(a) receiving the digital image;

(b) determining individually the sharpness, contrast, noise, and exposure of the digital image;

(c) determining a threshold individually for sharpness, contrast, noise, and exposure, which determined threshold or thresholds determines if the image is classified as a dud;

(d) determining an overall image quality parameter for the digital image by forming a linear combination of the sharpness, contrast, noise and exposure thresholds; and

(e) classifying the image as a dud based on the determination of the overall image quality parameter in step (d).

12. The computer program product as in claim 11, wherein the step of determining the sharpness includes the steps of:

obtaining an edge profile;

computing a histogram from the edge profile;

locating predetermined edges of the edge histogram; and

computing a centroid of gradients of the predetermined edges.

13. The computer program product as in claim 11, wherein the step of determining the contrast includes the steps of:

obtaining an edge profile;

computing an intensity histogram from the edge profile; and

calculating a standard deviation of the histogram.

14. The computer program product as in claim 11, wherein the step of determining the noise includes the steps of:

obtaining an edge profile;

computing a histogram from the edge profile; and

locating flat regions of the histogram; and

calculating a standard deviation of the flat regions.

15. The computer program product as in claim 11, wherein the step of determining the exposure includes the steps of:

comparing each image pixel to either an overexposure or underexposure threshold;

computing a percentage of pixels that are either above the overexposure threshold or below the underexposure threshold; and

determining the overexposure or underexposure based on the computed percentage.