METHOD AND APPARATUS FOR SEMI-AUTOMATIC FINGER EXTRACTION

• US 20140226856A1
• Filed: 02/06/2014
• Published: 08/14/2014
• Est. Priority Date: 02/06/2013
• Status: Active Grant

• Current Assignee: PFU Limited (Fujitsu Limited)
• Original Assignee: Fujitsu Limited
• Inventors: XIE, Shufu, HE, Yuan, SUN, Jun
• Granted Patent: US 9,311,538 B2
• Time in Patent Office: Days
• US Class Current: 382/103
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/155 Removing patterns interfering with the pattern to be recognised, such as ruled lines or underlines

Abstract

An image processing device includes: an inputting unit for performing a click on an object image contained in an image to obtain a clicked point; a calculating unit for calculating an edge map of the image; an estimating unit for estimating a color model of the object image based on the clicked point and the edge map; an object classifying unit for classifying each pixel in the image, based on the edge map and the color model, so as to obtain a binary image of the image; and a detecting unit for detecting a region containing the object image based on the binary image. The image processing device and method according to the present disclosure can improve the accuracy of detecting the boundary of an object image such as a finger image, thus facilitating removal of the object image from the image and making the processed image more nice-looking.

Claims

1. An image processing device comprising:

an inputting unit for performing a click on an object image contained in an image to obtain a clicked point;

a calculating unit for calculating an edge map of the image;

an estimating unit for estimating a color model of the object image based on the clicked point and the edge map;

an object classifying unit for classifying each pixel in the image, based on the edge map and the color model, so as to obtain a binary image of the image; and

a detecting unit for detecting a region containing the object image based on the binary image.

2. The device according to claim 1, wherein the calculating unit comprises:

a distance calculating unit for calculating a distance between a color of each pixel in the image and a color of the clicked point to obtain a distance map;

a distance gradient calculating unit for applying a gradient operator to the distance map to obtain a distance gradient image; and

an edge classifying unit for classifying a pixel having a distance gradient larger than a predetermined distance gradient threshold in the image into an edge pixel, and the other pixel in the image into a non-edge pixel.

3. The device according to claim 1, wherein the calculating unit comprises:

a distance calculating unit for calculating a distance between a color of each pixel in the image and a color of the clicked point to obtain a distance map;

a distance gradient calculating unit for applying a gradient operator to the distance map to obtain a distance gradient image;

a gray converting unit for converting the image from a color image to a gray image;

an intensity gradient calculating unit for applying a gradient operator to the gray image to obtain an intensity gradient image; and

an edge classifying unit for classifying a pixel having a distance gradient larger than a predetermined distance gradient threshold or having an intensity gradient larger than a predetermined intensity gradient threshold in the image into an edge pixel, and the other pixel in the image into a non-edge pixel.

4. The device according to claim 1, wherein the estimating unit comprises:

an extension region acquiring unit for acquiring an extension region containing the clicked point based on the clicked point and the edge map, the extension region being within the object image; and

a color model acquiring unit for acquiring the color model of the object image based on a color of each pixel within the extension region.

5. The device according to claim 4, wherein the extension region acquiring unit comprises:

a setting unit for setting a maximum extension region containing the clicked point;

a first searching unit for searching, as a left boundary pixel of the extension region, the first one of boundary pixels leftward in a horizontal direction from the clicked point;

a second searching unit for searching, as a right boundary pixel of the extension region, the first one of the boundary pixels rightward in the horizontal direction from the clicked point;

a third searching unit for searching, for each reference pixel between the left boundary pixel and the right boundary pixel in the horizontal direction, as an upper boundary pixel of the extension region, the first one of the boundary pixels upward in a vertical direction from the reference pixel; and

a fourth searching unit for searching, as an lower boundary pixel of the extension region, the first one of the boundary pixels downward in the vertical direction from the reference pixel, wherein,

the extension region acquiring unit sets a sliding window taking each pixel within the maximum extension region as a center, counts the number of edge pixels in the sliding window, and defines a pixel satisfying a condition that the number of the edge pixels in the sliding window is larger than a predetermined threshold as the boundary pixel.

6. The device according to claim 1, wherein the object classifying unit classifies a pixel in the image which is a non-edge pixel in the edge map and a distance from the color model is less than a color threshold into an object pixel, and the other pixel in the image into a non-object pixel.

7. The device according to claim 1, wherein the detecting unit comprises a noise removing unit for removing noise component in the binary image.

8. The device according to claim 7, wherein the noise removing unit comprises:

a connected component analyzing unit for performing connected component analysis algorithm on the binary image, so as to obtain a connected component in the binary image, each of pixels in the connected component being an object pixel; and

a pixel converting unit for converting each object pixel in the connected component into a non-object pixel if the connected component satisfies any of conditions of;

an area of the connected component being less than a predetermined area;

an aspect ratio of the connected component being larger than a predetermined ratio;

the object image being on the left side of the image and a distance between a left boundary of the connected component and a left boundary of the image being larger than a predetermined threshold;

or

the object image being on the right side of the image and a distance between a right boundary of the connected component and a right boundary of the image being larger than the predetermined threshold.

9. The device according to claim 8, wherein the detecting unit further comprises:

a connected component processing unit for acquiring a connected component where the clicked point locates according to the clicked point and searching a nearby connected component in a vertical direction; and

a filling unit for performing filling operation on the connected component containing the clicked point and the searched connected component, so as to obtain the filled connected component.

10. The device according to claim 9, wherein the detecting unit further comprises:

an expanding unit for performing an expanding operation on the filled connected component in the binary image.

11. The device according to claim 1, wherein the object image is a finger image.

12. The device according to claim 1, wherein the color of the clicked point is the color of a pixel at the clicked point, or an average color of pixels within a predetermined region containing the clicked point.

13. An image processing method comprising:

performing a click on an object image contained in an image to obtain a clicked point;

calculating an edge map of the image;

estimating a color model of the object image based on the clicked point and the edge map;

classifying each pixel in the image, based on the edge map and the color model, so as to obtain a binary image of the image; and

detecting a region containing the object image based on the binary image.

14. The method according to claim 13, wherein the step of calculating an edge map of the image comprises:

calculating a distance between a color of each pixel in the image and a color of the clicked point to obtain a distance map;

applying a gradient operator to the distance map to obtain a distance gradient image; and

classifying a pixel having a distance gradient larger than a predetermined distance gradient threshold in the image into an edge pixel, and the other pixel in the image into a non-edge pixel.

15. The method according to claim 13, wherein the step of calculating an edge map of the image comprises:

calculating a distance between a color of each pixel in the image and a color of the clicked point to obtain a distance map;

applying a gradient operator to the distance map to obtain a distance gradient image;

converting the image from a color image to a gray image;

applying a gradient operator to the gray image to obtain an intensity gradient image; and

classifying a pixel having a distance gradient larger than a predetermined distance gradient threshold or having an intensity gradient larger than a predetermined intensity gradient threshold in the image into an edge pixel, and the other pixel in the image into a non-edge pixel.

16. The method according to claim 13, wherein the step of estimating a color model of the object image based on the clicked point and the edge map comprises:

acquiring an extension region containing the clicked point based on the clicked point and the edge map, the extension region being within the object image; and

acquiring the color model of the object image based on a color of each pixel within the extension region.

17. The method according to claim 16, wherein the step of acquiring an extension region containing the clicked point based on the clicked point and the edge map comprises:

setting a maximum extension region containing the clicked point;

searching, as a left boundary pixel of the extension region, the first one of boundary pixels leftward in a horizontal direction from the clicked point;

searching, as a right boundary pixel of the extension region, the first one of the boundary pixels rightward in the horizontal direction from the clicked point; and

setting, for each reference pixel between the left boundary pixel and the right boundary pixel in the horizontal direction, an upper boundary pixel and a lower boundary pixel of the extension region by steps of;

searching, as the upper boundary pixel of the extension region, the first one of the boundary pixels upward in a vertical direction from the reference pixel; and

searching, as the lower boundary pixel of the extension region, the first one of the boundary pixels downward in the vertical direction from the reference pixel, wherein,

a sliding window is set taking each pixel within the maximum extension region as a center, the number of edge pixels in the sliding window is counted, and a pixel satisfying a condition that the number of the edge pixels in the sliding window is larger than a predetermined threshold is defined as the boundary pixel.

18. The method according to claim 13, wherein the step of classifying each pixel in the image based on the edge map and the color model so as to obtain a binary image of the image comprises:

classifying a pixel in the image which is a non-edge pixel in the edge map and a distance from the color model is less than a color threshold into an object pixel, and the other pixel in the image into a non-object pixel.

19. A program product comprising a machine-readable instruction code stored therein, wherein the instruction code, when read and executed by a computer, enables the computer to execute the method according to claim 13.