Mapping defects or dirt dynamically affecting an image acquisition device

• US 6,035,072 A
• Filed: 12/08/1997
• Issued: 03/07/2000
• Est. Priority Date: 12/08/1997
• Status: Expired due to Term

• Current Assignee: FotoNation Limited (Adeia, Inc.)
• Original Assignee: Robert Lee Read
• Inventors: Read, Robert Lee
• Primary Examiner(s): ROGERS, SCOTT A
• Application Number:
  US08/986,925
• Time in Patent Office: 820 Days
• Field of Search: 382/275, 382/260, 382/261, 382/254, 382/149, 382/141, 358/463, 358/448, 358/474, 358/406, 358/496, 358/483, 358/482, 348/125, 348/241, 348/246, 348/248, 348/249, 348/250, 348/251
• US Class Current: 382/275
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/164 Noise filtering
• CPC Class Codes: G06V 30/18095 Summing image-intensity values; Projection and histogram analysis
• CPC Class Codes: H04N 1/4097 Removing errors due external factors, e.g. dust, scratches

Abstract

Defects such as dirt, dust, scratches, blemishes, pits, or defective elements or pixels in a CCD, scanner, photocopier, or image acquiring device are dynamically detected by processing a plurality of images via a computer. A pristine object of calibration is not required. Stationary components of the video images are found and detected so as to produce a low false alarm probability. Text segmentation and measurement of total deviation based on variability related to high-frequency components of the video image are employed to prevent applying the process or method to printed text or graphics. Additional techniques optionally employed are median filtering, sample area detection, and dynamic adjustment of scores. In special cases, only moderately blank documents are used. The dynamic defect detection allows defect compensation, defect correction, and alerting the operator of defects.

Claims

1. A method of dynamically detecting defective regions of an image acquisition device that are corrupted by dirt or defects positionally associated with said image acquisition device comprising the steps of:

a) initializing a score target value h to a predetermined number,

b) initializing a map of potential defects P that associates with each defect the number of the image in which it was found,

c) acquiring a serially numbered image numbered q from said image acquisition device,

d) constructing a map of deviant regions from image number q that includes just those regions having at least a predetermined deviation of intensity or color from the surrounding area in said image,

e) adding each deviant region from said map of deviant regions and its image number q to P.

f) scoring each region r in said map of potential defects with a scoring function that is monotonic in positional correspondence between r and other regions P having image numbers differing from q,

g) removing from P regions obsolete relative to said scoring function,

h) detecting as defective regions just those regions that scored at least said score target value h,

i) repeating steps c through h;

whereby the operator may be alerted to the existence and position of defective regions, or said defective regions may be removed from or compensated for in the acquired images.

2. The method of claim 1 further including the step of qualifying said image numbered q by determining if substantial difference exists between said image numbered q and the most recent previously qualified image and wherein only qualified images are processed further.

3. The method of claim 1 further including the step of qualifying said image numbered q image by receiving a signal from said image acquisition device and wherein only qualified images are processed further.

4. The method of claim 1 further including the step of excluding non-sample areas from said image numbered q by means for discriminating sample-areas.

5. The method of claim 1 further including the step of designating as inactive segments of the image that contain text or graphics found by means for text segmenting and excluding inactive segments from further processing.

6. The method of claim 1 further including the step of designating as inactive those segments having at least a predetermined total deviation and excluding inactive segments from further processing.

7. The method of claim 1 further including the step of adjusting said score target value h based on the total deviation to attain a predetermined target false alarm.

8. The method of claim 1 wherein said deviant regions are limited to individual pixels.

9. The method of claim 1 further including the step of combining intersecting defective regions into non-intersecting regions before being reported.

10. The method of claim 9 wherein the scoring function counts the number of images out of the most recent n that have active segments positionally corresponding to r and that have a deviant region positionally corresponding to r, and n is a predetermined number.

11. The method of claim 10 further including the steps of:

a) qualifying said image numbered q by determining if substantial difference exists between said image numbered q and the most recent previously qualified image and wherein only qualified images are processed further,

b) excluding non-sample areas from said image numbered q by means for discriminating sample-areas,

c) designating as inactive those segments having at least a predetermined total deviation and excluding inactive segments from further processing.

12. A program storage device readable by a machine, tangibly embodying a program of instructions executable by the machine to perform method steps for dynamically detecting defective regions of an image acquisition device that are corrupted by dirt or defects positionally associated with said image acquisition device comprising the steps of:

a) initializing a score target value h to a predetermined number,

b) initializing a map of potential defects P that associates with each defect the number of the image in which it was found,

c) acquiring a serially numbered image numbered q from said image acquisition device,

d) constructing a map of deviant regions from image number q that includes just those regions having at least a predetermined deviation of intensity or color from the surrounding area in said image,

e) adding each deviant region from said map of deviant regions and its image number q to P,

f) scoring each region r in said map of potential defects with a scoring function that is monotonic in positional correspondence between r and other regions P having image numbers differing from q,

g) removing from P regions obsolete relative to said scoring function,

h) detecting as defective regions just those regions that scored at least said score target value h,

i) repeating steps c through h;

whereby the operator may be alerted to the existence and position of defective regions, or said defective regions may be removed from or compensated for in the acquired images.

13. The program storage device of claim 12 further including the step of qualifying said image numbered q by determining if substantial difference exists between said image numbered q and the most recent previously qualified image and wherein only qualified images are processed further.

14. The program storage device of claim 12 further including the step of excluding non-sample areas from said image numbered q by means for discriminating sample-areas.

15. The program storage device of claim 12 further including the step of designating as inactive segments of the image that contain text or graphics found by means for text segmenting and excluding inactive segments from further processing.

16. The program storage device of claim 12 further including the step of designating as inactive segments having at least a predetermined total deviation and excluding inactive segments from further processing.

17. The program storage device of claim 12 further including the step of combining intersecting defective regions into non-intersecting regions before being reported.

18. The program storage device of claim 17 wherein the scoring function counts the number of images out of the most recent n that have active segments positionally corresponding to r and that have a deviant region positionally corresponding to r, and n is a predetermined number.

19. The program storage device of claim 18 further including the steps of:

a) qualifying said image numbered q by determining if substantial difference exists between said image numbered q and the most recent previously qualified image and wherein only qualified images are processed further,

b) excluding non-sample areas from said image numbered q by means for discriminating sample-areas,

c) designating as inactive segments having at least a predetermined total deviation and excluding inactive segments from further processing.

20. An apparatus dynamically detecting defective regions of an image acquisition device that are corrupted by dirt or defects positionally associated with said image acquisition device comprising:

a) means for initializing a score target value h to a predetermined number,

b) means for initializing a map of potential defects P that associates with each defect the number of the image in which it was found,

c) means for acquiring a serially numbered image numbered q from said image acquisition device,

d) means for constructing a map of deviant regions from image number q that includes just those regions having at least a predetermined deviation of intensity or color from the surrounding area in said image,

e) means for adding each deviant region from said map of deviant regions and its image number q to P.

f) means for scoring each region r in said map of potential defects with a scoring function that is monotonic in positional correspondence between r and other regions P having image numbers differing from q,

g) means for removing from P regions obsolete relative to said scoring function,

h) means for detecting as defective regions just those regions that scored at least said score target value h,

whereby the operator may be alerted to the existence and position of defective regions, or said defective regions may be removed from or compensated for in the acquired images.