Removing inserted text from an image using extrapolation for replacement pixels after optical character recognition

1. A method of removing inserted text from a digital image comprising, with a computer system:

recognizing said inserted text in said digital image using optical character recognition;

replacing pixels of said digital image corresponding to said inserted text so as to remove said inserted text from said digital image; and,

after said replacing pixels corresponding to said inserted text, blending replacement pixels with adjacent pixels;

wherein recognizing said inserted text further comprises presenting presumed inserted text to a user for validation, and

in which replacing pixels of said digital image corresponding to said inserted text so as to remove said inserted text from said digital image comprises determining a brightness value for the replacement pixels by separately averaging brightness values of pixels in each layer of pixels of a pixel square, each layer of pixels in said pixel square being defined by distance from said replacement pixel.

2. A method of removing inserted text from a digital image comprising:

with a data processing system having an interface for receiving said digital image from an input device, recognizing said inserted text in said digital image using optical character recognition; and

with said data processing system, electronically replacing pixels of said digital image corresponding to said inserted text so as to remove said inserted text from said digital image;

wherein replacing said pixels comprises extrapolating values for replacement pixels from surrounding pixels of said digital image;

wherein said extrapolating comprises blending values from a pixel square, comprised of multiple pixels, located to one side of an inserted text character pixel being replaced to generate a value for a replacement pixel, and

in which replacing said pixels comprises determining a brightness value for a replacement pixel by separately averaging brightness values of pixels in each layer of pixels of said pixel square, each layer of pixels in said pixel square being defined by distance from said replacement pixel.

3. The method of claim 2 , further comprising determining a color for the replacement pixel using the same pixel square by blending color values for said pixels of said pixel square and weighting color values from different layers of pixels differently in the blend of color values.

4. The method of claim 2 , further comprising repeating said extrapolating for each pixel to be replaced, starting with pixels at an outer edge of said inserted text and moving inward.

5. A method of removing inserted text from a digital image comprising:

with a data processing system having an interface for receiving said digital image from an input device, recognizing said inserted text in said digital image using optical character recognition; and

with said data processing system, electronically replacing pixels of said digital image corresponding to said inserted text so as to remove said inserted text from said digital image;

wherein replacing said pixels comprises extrapolating values for replacement pixels from surrounding pixels of said digital image;

wherein said extrapolating comprises blending values from a pixel square located entirely to one side of an inserted text character pixel being replaced to generate a value for a replacement pixel;

wherein said blending comprises averaging values of pixels from said pixel square in different layers, layers being defined by distance within the pixel square from the inserted text character pixel being replaced.

6. The method of claim 5 , wherein an average value from said different layers is assigned a different weight based on proximity of the corresponding layer to said inserted text character pixel being replaced.

7. A method of removing inserted text from a digital image comprising;

with a data processing system having an interface for receiving said digital image from an input device, recognizing said inserted text in said digital image using optical character recognition; and

with a data processing system, electronically replacing pixels of said digital image corresponding to said inserted text so as to remove said inserted text from said digital image;

wherein replacing said pixels comprises extrapolating values for replacement pixels from surrounding pixels of said digital image;

wherein said extrapolating comprises blending values from a pixel square located to one side of an inserted text character pixel being replaced to generate a value for a replacement pixel; and

additionally averaging values of pixels from a larger square located to one side of said inserted text character pixel being replaced to generate a value for said replacement pixel.

8. A computer program product for removing inserted text from a digital image comprises an inserted text removal program stored on a non-transitory computer-readable medium, said program comprising:

an optical character recognition module for recognizing an inserted text in a digital image; and

an extrapolation module for replacing pixels corresponding to said inserted text in said digital image with replacement image pixels so as to remove said inserted text from said digital image;

wherein said extrapolation module blends values from a pixel square, comprised of multiple pixels, located entirely to one side of an inserted text character pixel being replaced to generate a value for a replacement pixel, and

in which said extrapolation module determines a brightness value for a replacement pixel by separately averaging brightness values of pixels in each layer of pixels of said pixel square, each layer of pixels in said pixel square being defined by distance from said replacement pixel.

9. The product of claim 8 , wherein the extrapolation module, after said replacing pixels corresponding to said inserted text, blends replacement pixels with adjacent pixels.

10. The product of claim 8 , wherein said optical character recognition module compares recognized characters with known date formats to identify a date stamp.

11. The product of claim 8 , wherein said program further presents presumed inserted text to a user for validation.

12. A photo printing kiosk comprising a data processing system, said data processing system comprising:

an electronic interface for receiving a digital image from an input device;

an optical character recognition module for recognizing inserted text in said digital image; and

an extrapolation module for replacing pixels corresponding to said inserted text in said digital image with replacement image pixels so as to remove said inserted text from said digital image;

wherein said extrapolation module blends values from a pixel square located to one side of an inserted text character pixel being replaced to generate a value for a replacement pixel;

wherein said blending comprises combining values of pixels from said square in different layers, layers being defined by distance within the square from the pixel being replaced.

13. The kiosk of claim 12 , wherein said optical character recognition module compares recognized characters with known date formats or image labels to identify said inserted text.

14. The kiosk of claim 12 , further comprising a user interface, wherein said optical character recognition module is configured to present presumed inserted text to a user for validation using said user interface.

15. The kiosk of claim 12 , wherein said extrapolation module extrapolates values for said replacement image pixels from pixels surrounding said pixels corresponding to said inserted text.

16. The kiosk of claim 12 , further comprising an electronic display for receiving said digital image with said inserted text removed.

17. The kiosk of claim 12 , further comprising a printer for receiving said digital image with said inserted text removed.