Methods and systems for auto correction of defects in a scanned image

Patent metadata

Field	Value
Publication number	US-9998628-B1
Application number	US-201715477145-A
Country	US
Kind code	B1
Filing date	Apr 3, 2017
Priority date	Apr 3, 2017
Publication date	Jun 12, 2018
Grant date	Jun 12, 2018

Abstract

Disclosed are methods and multi-function devices for auto-correction of shadow effects in a scanned image. The method includes receiving a first image including a colored twin-page scanned document including at least one of a shadow region in a central part of the first image, at least one darker region caused at one or more borders of the first image, and distorted orientation with respect to a base of the first image; re-positioning the shadow region present in the central part to generate a re-positioned first binary image such that the shadow region becomes perpendicular to the base; detecting a shadow region in a central region of the re-positioned first binary image; removing the detected at least one shadow region from the re-positioned first binary image to generate a second image; and correcting a skewed orientation of each of a cropped left side and right side of the second image.

First claim

What is claimed is:

1. A method for automatically correcting a plurality of shadow effects present in a scanned image, comprising:
receiving, by a transceiving device, a first image comprising a colored twin-page scanned document, wherein the first image comprising at least one of a shadow region in a central part of the first image such that the shadow region is not perpendicular to the central part of the first image, at least one darker region caused at one or more borders of the first image, and distorted orientation with respect to a base of the first image;
re-positioning, by an image processing device, the shadow region present in the central part of the first image to generate a re-positioned first binary image such that the shadow region becomes perpendicular to the base of the first image;
detecting, by the image processing device, at least one shadow region in a central region of the re-positioned first binary image;
removing, by the image processing device, the detected at least one shadow region from the re-positioned first binary image to generate a second image comprising a left side and a right side; and
correcting, by the image processing device, a skewed orientation of each of a cropped left side and a cropped right side of the second image.

2. The method of claim 1 , wherein re-positioning the shadow region present in the central part of the first image further comprises:
converting, by the image processing device, the first image into a first grey scale image, wherein the first grey scale image comprises an eight-bit grey scale image;
converting, by the image processing device, the first grey scale image into a first binary image comprising a plurality of black pixels based on a grey threshold, the grey threshold depends on a nature of the first grey scale image;
removing, by the image processing device, content including text content from the first binary image by replacing the plurality of black pixels with a plurality of white pixels, wherein a flood fill operation is applied to remove the text content;
applying, by the image processing device, a dilation method for removing sporadic noise segments to generate a dilated image;
extracting, by the image processing device, one or more edges from the dilated image to generate an edge image;
determining, by the image processing device, a degree of rotation “θ” with respect to the base of the first grey scale image; and
rotating, by the image processing device, the first image based on the degree of rotation “θ” to generate the re-positioned first binary image, wherein a maximum returned degree of rotation “θ” is considered for rotating the first image and making the shadow region perpendicular to the base.

3. The method of claim 2 , wherein the degree of rotation “θ” is determined by applying a Hough line transformation method.

4. The method of claim 2 , wherein correcting the skewed orientation of the second image further comprising:
for each of the cropped left side and the cropped right side of the second image:
converting a cropped side of the cropped left side and the cropped right side of the second image into a second grey scale image comprising one or more black pixels, wherein the cropped left side and the cropped right side have different skewed orientation;
removing text content from the second grey scale image by replacing the one or more black pixels with one or more white pixels, wherein a flood fill operation is applied for removing the text content;
generate an eroded image by expanding an area of the second grey scale image by applying an image erosion method;
creating a complementary image corresponding to the eroded image;
generate a painted eroded image replacing a pixel at a position in the eroded image with a pixel value present in a same position in the second grey scale image when a pixel other than a white pixel is encountered in the complementary image;
creating a second binary image of the painted eroded image;
extracting one or more edges from the second binary image; and
rotating the second binary image to remove a skew from the second binary image.

5. The method of claim 4 , wherein the second binary image is rotated for removing the skew by applying a Hough line transformation method, wherein a maximum returned degree is considered for rotating the second binary image.

6. The method of claim 4 , wherein the one or more edges are extracted from the second binary image by applying a canny operator.

7. A multi-function device for automatically correcting a plurality of shadow effects present in a scanned image, comprising:
a transceiving device configured to receive a first image comprising a colored twin-page scanned document, wherein the first image comprising at least one of a shadow region in a central part of the first image such that the shadow region is not perpendicular to the central part of the first image, at least one darker region caused at one or more borders of the first image, and distorted orientation with respect to a base of the first image; and
an image processing device configured to:
re-position the shadow region in the central part of the first image to generate a re-positioned first binary image such that the shadow region becomes perpendicular to the base of the first image;
detect at least one shadow region in a central region of the re-positioned first binary image;
remove the detected at least one shadow region from the re-positioned first binary image to generate a second image comprising a left side and a right side; and
correct a skewed orientation of each of a cropped left side and a cropped right side of the second image.

8. The multi-function device of claim 7 , wherein the image processing device is further configured to:
convert the first image into a first grey scale image, wherein the first grey scale image is an eight-bit grey scale image;
convert the first grey scale image into a first binary image comprising a plurality of black pixels based on a grey threshold, the grey threshold depends on a nature of the first grey scale image;
remove content including text content from the first binary image by replacing the plurality of black pixels with a plurality of white pixels, wherein a flood fill operation is applied to remove the text content;
apply a dilation method for removing sporadic noise segments to generate a dilated image;
extract one or more edges from the dilated image to generate an edge image;
determine a degree of rotation “θ” with respect to the base of the first grey scale image, wherein the degree of rotation “θ” is determined by applying a Hough line transformation method; and
rotate the first image based on the degree of rotation “θ” to generate the re-positioned first binary image, wherein a maximum returned degree of rotation “θ” is considered for rotating the first image and making the shadow region perpendicular to the base.

9. The multi-function device of claim 7 , wherein for each of the cropped left side and the cropped right side of the second image, the image processing device is configured to:
convert the second image into a second grey scale image comprising one or more black pixels;
remove text content from the second grey scale image by replacing the one or more black pixels with one or more white pixels, wherein a flood fill operation is applied for removing the text content;
expand an area of the second grey scale image by applying an image erosion method to generate an eroded image;
create a complementary image corresponding to the eroded image;
generate a painted eroded image by replacing a pixel at a position in the eroded image with a pixel value present in a same position in the second grey scale image when a pixel oth

Assignees

• Xerox Corp

Inventors

• Palaniyappan Mohan
• Puttuswamy Punitha
• Das Arindam

Classifications

• H04N1/4097 Primary

  Removing errors due external factors, e.g. dust, scratches · CPC title

• H04N1/4092

  Edge or detail enhancement · CPC title

• G06T5/006

  Physics · mapped topic

• G06T2207/10008

  from scanner, fax or copier · CPC title

• H04N1/387

  Composing, repositioning or otherwise {geometrically} modifying originals · CPC title