Binarization of an image

• US 20070025625A1
• Filed: 07/29/2005
• Published: 02/01/2007
• Est. Priority Date: 07/29/2005
• Status: Active Grant

• Current Assignee: Nokia Technologies Oy (Nokia Corporation)
• Original Assignee: Nokia Corporation
• Inventors: Vehvilainen, Markku, Burian, Adrian
• Granted Patent: US 7,636,467 B2
• Time in Patent Office: Days
• US Class Current: 382/237
• CPC Class Codes: G06T 2207/20012 Locally adaptive
• CPC Class Codes: G06T 2207/20021 Dividing image into blocks, subimages or windows
• CPC Class Codes: G06T 2207/30176 Document
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/136 involving thresholding
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/162 Quantising the image signal

Abstract

For binarizing an image, which is composed of pixels, the image is split into two semi-images. Local parameters are initialized based on values of pixels that lie in an area adjacent to a boundary separating the semi-images. A binarization is then performed separately for each of the semi-images using an adaptive threshold, wherein the adaptive threshold is calculated for each of the semi-images proceeding from the initialized local parameters.

Claims

1. A method for binarizing an image, which image is composed of pixels, said method comprising:

splitting said image into two semi-images;

initializing local parameters based on values of pixels that lie in an area adjacent to a boundary separating said semi-images; and

performing a binarization separately for each of said semi-images using an adaptive threshold, wherein said adaptive threshold is calculated for each of said semi-images proceeding from said initialized local parameters.

2. The method according to claim 1, wherein said pixels are arranged in columns and rows, wherein said boundary is a vertical line, wherein said initialized local parameters comprise for each row an initial sum of pixel values in said row in said area adjacent to said vertical line, and wherein said adaptive threshold is adapted for each pixel in each row of each semi-image based on a windowed moving sum proceeding from an initial sum associated to a respective row.

3. The method according to claim 2, further comprising splitting said image along a horizontal line into two further semi-images, initializing further local parameters, said further initialized local parameters comprising for each column an initial sum of pixel values in an area adjacent to said horizontal line, and correcting said binarized image separately in each of said further semi-images using said further initialized local parameters.

4. The method according to claim 2, wherein a windowed moving sum which is determined for a particular pixel is averaged with a windowed moving sum which is determined for a pixel adjacent to said particular pixel, before said windowed moving sum is used for calculating an adaptive threshold for said particular pixel.

5. The method according to claim 1, wherein said pixels are arranged in columns and rows, wherein said boundary is a horizontal line, wherein said initialized local parameters comprise for each column an initial sum of pixel values in said column in said area adjacent to said horizontal line, and wherein said adaptive threshold is calculated for each pixel in each column of each semi-image based on a windowed moving sum proceeding from an initial sum associated to a respective column.

6. The method according to claim 5, further comprising splitting said image along a vertical line into two further semi-images, initializing further local parameters, said further initialized local parameters comprising for each row an initial sum of pixel values in an area adjacent to said vertical line, and correcting said binarized image separately in each of said further semi-images using said further initialized local parameters.

7. The method according to claim 4, wherein a windowed moving sum which is determined for a particular pixel is averaged with a windowed moving sum which is determined for a pixel adjacent to said particular pixel, before said windowed moving sum is used for calculating an adaptive threshold for said particular pixel.

8. The method according to claim 1, further comprising a step of computing global parameters for said image, wherein said adaptive threshold is calculated for each of said semi-images taking in addition account of said global parameters.

9. The method according to claim 8, wherein said global parameters are determined separately for each of said semi-images.

10. The method according to claim 8, wherein said initialized local parameters comprise initial sums of pixel values in said area adjacent to said boundary, wherein adaptive thresholds are calculated for pixels of each semi-image based on windowed moving sums of pixel values proceeding from a respective initial sum, and wherein an adaptive threshold for a particular pixel is determined by applying the following operations to a moving sum for the particular pixel:

a division by a number of pixels considered in said moving sum to obtain a moving average;

a multiplication with a value of an adjustment function for said moving average, wherein said adjustment function has a decreasing value in a first range of moving averages, a constant value in a second range of moving averages and a further decreasing value in a third range of moving averages, said ranges being defined by a first and a second one of said computed global parameter values; and

a multiplication with third computed global parameter value that lies between said first and said second computed global parameter values.

11. The method according to claim 1, wherein said image is a color image, said method comprising a preceding step of converting said color image into a grayscale image.

12. The method according to claim 1, further comprising applying binary corrections to said binarized image.

13. The method according to claim 12, wherein said binary corrections comprise at least one of a binary media filtering and a binary morphological filtering.

14. The method according to claim 1, in which a parallel processing is carried out in accordance with at least one of the following:

said initializing of local parameters is carried out in parallel with a computing of global parameters for said image;

within said initializing of local parameters for an image, in which said pixels are arranged in columns and rows, a separate initializing of local parameters is carried out in parallel for respectively at least two columns or at least two rows;

said initializing of local parameters in one direction is carried out in parallel with an additional initializing of local parameters in another direction;

within said binarization a calculation of an adaptive threshold is carried out in parallel for each of said semi-images; and

within said binarization for an image, in which said pixels are arranged in columns and rows, a separate calculation of an adaptive threshold is carried out in parallel for respectively at least two columns or at least two rows in a respective one of said semi-images.

15. An image processing chipset, which comprises for a binarization of an image composed of pixels:

an initializing portion adapted to split an image into two semi-images and to initialize local parameters based on values of pixels that lie in an area adjacent to a boundary separating said semi-images; and

a binarizing portion adapted to perform a binarization separately for each of two semi-images of an image using an adaptive threshold, wherein said adaptive threshold is calculated for each of said semi-images proceeding from local parameters initialized by said initializing portion.

16. The image processing chipset according to claim 15, further comprising a global parameter extraction portion adapted to compute global parameters for said image, wherein said binarizing portion is adapted to calculate said adaptive threshold for each of said semi-images taking account of said global parameters also.

17. An electronic device, which comprises for a binarization of an image composed of pixels:

an initializing portion adapted to split an image into two semi-images and to initialize local parameters based on values of pixels that lie in an area adjacent to a boundary separating said semi-images; and

a binarizing portion adapted to perform a binarization separately for each of two semi-images of an image using an adaptive threshold, wherein said adaptive threshold is calculated for each of said semi-images proceeding from local parameters initialized by said initializing portion.

18. The electronic device according to claim 17, further comprising a global parameter extraction portion adapted to compute global parameters for said image, wherein said binarizing portion is adapted to calculate said adaptive threshold for each of said semi-images taking account of said global parameters also.

19. The electronic device according to claim 17, wherein said electronic device is one of a mobile phone, a computer, a digital camera, a personal digital assistant, and a portable device comprising a camera functionality.

20. A software program product, in which a software code for binarizing an image is stored, wherein said image is composed of pixels, said software code realizing the following steps when being executed by a processing unit of an electronic device:

splitting said image into two semi-images;

initializing local parameters based on values of pixels that lie in an area adjacent to a boundary separating said semi-images; and

performing a binarization separately for each of said semi-images using an adaptive threshold, wherein said adaptive threshold is calculated for each of said semi-images proceeding from said initialized local parameters.

21. The software program product according to claim 20, wherein when being executed by a processing unit of an electronic device said software code further realizes a step of computing global parameters for said image, said adaptive threshold being calculated for each of said semi-images taking account of said global parameters also.