Extended dynamic range imaging system

• US 7,583,305 B2
• Filed: 07/07/2004
• Issued: 09/01/2009
• Est. Priority Date: 07/07/2004
• Status: Active Grant

• Current Assignee: Omnivision Technologies Incorporated (Will Semiconductor Co., Ltd. Shanghai)
• Original Assignee: Eastman Kodak Company
• Inventors: McDermott, Bruce C.
• Primary Examiner(s): Ometz; David L
• Assistant Examiner(s): Chon; Peter
• Application Number:
  US10/887,071
• Publication Number:
  US 20060007498A1
• Time in Patent Office: 1,882 Days
• Field of Search: 348/229.1, 348/362, 348/297, 348/296, 348/220.1, 348/221.1, 348/222.1, 348/223.1, 348/224.1, 348/230.1, 348/294, 348/295, 348/298, 348/345, 348/349, 348/353, 348/354, 348/355, 348/356, 348/363, 348/364, 348/365, 348/302, 348/308, 358/443, 358/482, 358/208.1
• US Class Current: 348/297
• CPC Class Codes: H04N 23/71 Circuitry for evaluating the brightness variation
• CPC Class Codes: H04N 25/533 by using differing integration times for different sensor regions

Abstract

A method for decreasing integration time of saturated paxels within an imager; wherein the method includes decreasing the integration time of the saturated paxels within an imager according to scene data from a captured image.

Claims

1. A method for extending dynamic range of a captured image captured by an image sensor having a plurality of pixels, the method comprising the steps of:

capturing a first image using a first integration time for the plurality of pixels;

analyzing the first image on a paxel by paxel basis to locate one or more paxels having an incorrect intergration time, wherein each paxel comprises a group of pixels associated geographically with each other and the pixel values in the group of pixels for each paxel are averaged together to obtain a paxel value for each paxel;

determining an adjusted integration time for each paxel having an incorrect integration time, wherein a magnitude of an increment of adjustment for each integration time is selected to be one of at least two different increments based on a total number of paxels having incorrect integration times; and

capturing a second image using a respective adjusted integration time for each paxel and using the first integration time for the remaining pixels, wherein the respective adjusted integration times are different from the first integration time.

2. The method claimed in claim 1, further comprising the step of storing an X-Y address for each paxel within the captured image.

3. The method claimed in claim 1, wherein a large total number of paxels causes a large magnitude of adjustment to at least one respective adjusted integration time.

4. The method claimed in claim 1, wherein a small total number of paxels causes a small magnitude of adjustment to at least one respective adjusted integration time.

5. The method claimed in claim 1, further comprising the step of repeatedly capturing the second image until each respective adjusted integration time is a correct integration time for each group of pixels included in each paxel.

6. The method claimed in claim 1, further comprising the steps of:

determining whether each paxel value is equal to or greater than a constant, wherein the constant represents a value at which a paxel has an incorrect integration time; and

if one or more paxels has a paxel value that is equal to or greater than the constant, recording an X-Y address for each paxel having an incorrect integration time within the first image.

7. A method for extending a dynamic range of a captured image captured by an image sensor comprising a plurality of pixels, the method, comprising the steps of:

a) capturing a first image using a first integration time for the plurality of pixels;

b) analyzing the first image on a paxel by paxel basis to locate one or more paxels having an incorrect integration time, wherein each paxel comprises a group of pixels associated geographically with each other and the pixel values for the group of pixels in each paxel are averaged together to obtain a paxel value for each paxel;

c) determining whether each paxel value is equal to or greater than a first constant, wherein the first constant represents a value at which a paxel has an incorrect integration time;

d) if one or more paxels has a paxel value that is equal to or greater than the first constant, recording an X-Y address for each paxel having an incorrect integration time within the first image;

e) determining a total number of paxels having incorrect integration times in the first image;

f) comparing the total number of paxels having incorrect integration times to a second constant;

g) determining an adjusted integration time for each paxel having an incorrect integration time, wherein a magnitude of an increment of adjustment for each integration time is selected to be one of at least two different increments based on whether the total number of paxels having incorrect integration times is less than the second constant or greater than the second constant; and

h) capturing a second image and varying each respective integration time of the paxels having incorrect integration times while using the first integration time for the remaining pixels.

8. The method claimed in claim 7, wherein a large total number of paxels causes a large magnitude of adjustment to at least one respective adjusted integration time.

9. The method claimed in claim 7, wherein a small total number of paxels causes a small magnitude of adjustment to at least one respective adjusted integration time.