Whole order orientation method and apparatus

• US 5,642,443 A
• Filed: 10/12/1994
• Issued: 06/24/1997
• Est. Priority Date: 10/12/1994
• Status: Expired due to Term

• Current Assignee: Apple Inc.
• Original Assignee: Eastman Kodak Company
• Inventors: Goodwin, Robert Melvin, Manico, Joseph Anthony
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): SHALWALA, BIPIN H
• Application Number:
  US08/322,241
• Time in Patent Office: 986 Days
• Field of Search: 382/286, 382/289, 382/291, 382/293, 382/295, 382/296, 348/580-183
• US Class Current: 455/435.1
• CPC Class Codes: G06T 7/77 using statistical methods
• CPC Class Codes: G06V 10/242 by image rotation, e.g. by 90 degrees

Abstract

To determine the orientation of a set of recorded images, the recorded images are scanned. The scanning operation obtains information regarding at least one scene characteristic distributed asymmetrically in the separate recorded images. Probability estimates of orientation of each of the recorded images for which at least one scene characteristic is obtained are determined as a function of asymmetry in distribution of the scene characteristic. Probability of orientation for the set of recorded images is determined from the probability estimates of orientation of each of the recorded images in the set.

Claims

1. A method of determining orientation of a set of recorded images, comprising the steps of:

(a) scanning a set of recorded images, said set including a plurality of recorded images, to obtain information regarding at least one scene characteristic distributed asymmetrically in the separate recorded images;

(b) determining probability estimates of orientation of each of the recorded images for which at least one scene characteristic is obtained as a function of asymmetry in distribution of the scene characteristic; and

(c) determining a probability of orientation for said set of recorded images from the probability estimates of orientation of each of the recorded images in said set.

2. A method according to claim 1, further comprising the step of (d) image processing said set of recorded images according to the probability of orientation for said set of recorded images, determined in step (c).

3. A method according to claim 1, wherein step (c) of determining the probability of orientation for the set of recorded images is carried out using a Bayesian probability propagation model.

4. A method according to claim 1, wherein said at least one scene characteristic is color.

5. A method according to claim 1, wherein said at least one scene characteristic is modulation.

6. A method according to claim 1, wherein said step (a) of scanning comprises obtaining information regarding a plurality of scene characteristics.

7. A method according to claim 6, wherein said step (c) of determining probability of orientation for the set of recorded images is carried out employing information from different scene characteristics of the plurality of scene characteristics to determine said probability of orientation.

8. A method according to claim 7, wherein at least some of the plurality of scene characteristics are correlated to one another.

15. A method according to claim 1, further comprising the steps of:

(d) ascertaining a probability that a certain type of image recorder was used to create said set of recorded images; and

(e) utilizing said probability from said ascertaining step in step (c).

9. An apparatus for determining orientation of a set of individual recorded images, comprising:

a scanner which is operative to scan recorded images and provide digital signals representative of said recorded images; and

an orientation processor, coupled to said scanner to receive said digital signals, said processor being operative to execute the steps of;

determining the asymmetric distribution of at least one scene characteristic in said recorded images; and

determining the probability of orientation for said set of recorded images as a function of the determined asymmetric distribution of the at least one scene characteristic.

10. An apparatus according to claim 9, wherein said orientation processor is operative to determine asymmetric distribution by determining probability estimates of orientation of each of the individual recorded images, for which the at least one scene characteristic is obtained, as a function of the asymmetric distribution of said at least one scene characteristic in said recorded images.

11. An apparatus according to claim 10, wherein said orientation processor is operative to determine the probability of orientation for said set of recorded images includes in accordance with a Bayesian probability propagation model employing the determined probability estimates of orientation of each of the individual recorded images to produce said probability of orientation for said set of recorded images.

12. An apparatus according to claim 9, wherein said at least one scene characteristic includes color.

13. An apparatus according to claim 9, wherein said at least one scene characteristic includes modulation.

14. An apparatus according to claim 9, wherein said orientation processor is operative to determine said asymmetric distribution by determining the asymmetric distribution of a plurality of scene characteristics.