System, method, and apparatus for orienting images

• US 6,968,081 B1
• Filed: 11/14/2000
• Issued: 11/22/2005
• Est. Priority Date: 11/15/1999
• Status: Expired due to Fees

• Current Assignee: Luminys Systems Corp.
• Original Assignee: Luminys Systems Corp.
• Inventors: Judson, Ross, Meenan, Patrick
• Primary Examiner(s): AHMED, SAMIR ANWAR
• Application Number:
  US09/712,658
• Time in Patent Office: 1,834 Days
• Field of Search: 382/155, 382/156, 382/157, 382/158, 382/159, 128/925, 706/15, 706/16, 706/20, 706/25, 706/26, 706/27, 700/47, 700/48
• US Class Current: 382/157
• CPC Class Codes: G06T 7/74 involving reference images or patches
• CPC Class Codes: G06V 10/242 by image rotation, e.g. by 90 degrees

Abstract

The present invention is directed to a system, method, and apparatus for orienting images. A neural net is trained with images of known orientation and an indicator indicating such known orientation. Images of unknown orientation are then input to the neural net and the orientation is determined based on the output of the neural net.

Claims

1. A method for orienting an image, said method comprising:

training a neural net at least one image with known orientation associated with at least one indicator, wherein the at least one indicator is assigned a first value if the at least one image is properly oriented and a second value if the at least one image is improperly oriented;

inputting the image to the neural net;

receiving the output of the neural net, responsive to submitting the image; and

determining the orientation of the image responsive to receiving the output.

2. The method of claim 1, wherein submitting the image to the neural net comprises:

inputting the image to the neural net with a first orientation; and

inputting the image to the neural net with a second orientation.

3. The method of claim 2, wherein receiving the output of the neural net comprises:

receiving a first output of the neural net responsive to inputting the image to the neural net with the first orientation; and

receiving a second output of the neural net responsive to inputting the image to the neural net with the second orientation.

4. The method of claim 3, wherein determining the orientation of the image comprises:

selecting the first orientation wherein the first output is closer to the first value; and

selecting the second orientation wherein the second output is closer to the second value.

5. The method of claim 1, further comprising:

reducing the resolution of the image.

6. An article of manufacture comprising computer readable medium, said computer readable medium storing a plurality of executable instructions, said plurality of executable instructions comprising means for:

training a neural net at least one image with known orientation associated with at least one indicator, wherein the at least one indicator is assigned a first value if the at least one image is properly oriented and a second value if the at least one image is improperly oriented;

inputting the image to the neural net;

receiving the output of the neural net, responsive to submitting the image; and

determining the orientation of the image responsive to receiving the output.

7. The article of manufacture of claim 6, wherein the means for submitting the image to the neural net comprises means for:

inputting the image to the neural net with a first orientation; and

inputting the image to the neural net with a second orientation.

8. The article of manufacture of claim 7, wherein the means for receiving the output of the neural net comprises means for:

receiving a first output of the neural net responsive to inputting the image to the neural net with the first orientation; and

receiving a second output of the neural net responsive to inputting the image to the neural net with the second orientation.

9. The article of manufacture of claim 8, wherein the means for determining the orientation of the image comprises means for:

selecting the first orientation wherein the first output is closer to the first value; and

selecting the second orientation wherein the second output is closer to the second value.

10. The article of manufacture of claim 6, further comprising means for:

compressing the image.

11. A computer system for orienting a digital image, said computer system comprising:

a processor for performing processing functions;

memory;

means for training a neural net at least one image with known orientation associated with at least one indicator, wherein the at least one indicator is assigned a first value if the at least one image is properly oriented and a second value if the at least one image is improperly oriented;

means for inputting the image to the neural net;

means for receiving the output of the neural net, responsive to submitting the image; and

means for determining the orientation of the image responsive to receiving the output.

12. The computer system of claim 11, wherein the means for submitting the image to the neural net comprises means for:

inputting the image to the neural net with a first orientation; and

inputting the image to the neural net with a second orientation.

13. The computer system of claim 12, wherein the means for receiving the output of the neural net comprises means for:

receiving a first output of the neural net responsive to inputting the image to the neural net with the first orientation; and

receiving a second output of the neural net responsive to inputting the image to the neural net with the second orientation.

14. The computer system of claim 13, wherein the means for determining the orientation of the image comprises means for:

selecting the first orientation wherein the first output is closer to the first value; and

selecting the second orientation wherein the second output is closer to the second value.

15. The computer system of claim 11, further comprising means for:

compressing the image.