Image segmentation method with enhanced noise elimination function

• US 6,449,391 B1
• Filed: 09/09/1999
• Issued: 09/10/2002
• Est. Priority Date: 10/13/1998
• Status: Expired due to Term

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Ku, Jeong-Hoe
• Primary Examiner(s): Au, Amelia M.
• Assistant Examiner(s): Dastouri, Mehrdad
• Application Number:
  US09/392,396
• Time in Patent Office: 1,097 Days
• Field of Search: 382/174, 382/176, 382/177, 382/178, 382/187, 382/190, 382/192, 382/193, 382/194, 382/205, 382/266, 382/267, 382/102, 382/104, 382/199, 382/200, 382/105, 358/462, 358/463
• US Class Current: 382/194
• CPC Class Codes: G06T 2207/30248 Vehicle exterior or interior
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/162 Quantising the image signal

Abstract

Segmentation of a character-containing image, made by placing a dark mark against a light background, is accomplished by establishing a vertical pixel projection for each pixel column in the image. For more exact segmentation, the image is horizontally scanned in each row of the image after detecting edges of characters. The scanned results and thresholds are compared to verify whether edges and spaces of the scanned row are noise or discontinuous edges. The thresholds correlate with neighboring row data of the image. According to the results of the comparison, the discontinuous edges are connected and the noise is eliminated. Thereby, the character-containing image can be segmented into each character without segmentation errors.

Claims

1. A method for segmenting an image formed of a plurality of pixels, comprising the steps of:

detecting edges of the image, the edge-detected image including real edges and noise;

scanning rows of the image to discriminate between the real edges and the noise,

eliminating the noise of the image;

projecting the image; and

segmenting the image,

wherein the scanning step includes the steps of;

discriminating between real spaces and discontinuous edges;

connecting the discontinuous edges between two adjacent edges;

generating a first value by counting pixels corresponding to the real edges and the noise;

generating a second value by counting pixels corresponding to the real spaces and the discontinuous edges; and

comparing the first value and the second value with predetermined first threshold and second threshold, respectively, to verify the counted pixels according to the results of the comparison.

2. The method of claim 1, wherein the first and second thresholds limit ranges of the first and the second values, respectively, and correlate with neighboring data of the image.

3. The method of claim 1, wherein the verifying step includes verifying the counted pixels to the noise of the image when the first value is less than the first threshold and the second value is greater than the second threshold.

4. The method of claim 1, wherein the verifying step includes verifying the counted pixels to the discontinuous edges of the image when the first value is greater than the first threshold and the second value is less than the second threshold.

5. The method of claim 1, wherein the first threshold depends on a threshold of a previous row and an average number of pixels verified to edges in the previous row.

6. The method of claim 1, wherein the second threshold depends on a threshold of a previous row and an average number of pixels verified to spaces in the previous row.

7. The method of claim 1, wherein the first value is equal to a summation of pixels corresponding to edges of a row of an image.

8. The method of claim 1, wherein the second value is equal to a summation of pixels corresponding to spaces of a row of an image.

9. The method of claim 4, further including a step includes the step of connecting the discontinuous edges.

10. A method for segmenting an image formed of a plurality of pixels, comprising the steps of:

detecting edges and noise of each row of the image;

discriminating between the edges and the noise detected;

detecting spaces and discontinuous edges of each row of the image;

discriminating between the spaces and the discontinuous edges detected;

eliminating the noise discriminated from the edges;

connecting the discontinuous edges discriminated from the spaces;

projecting the image; and

segmenting the image,

wherein the step of discriminating includes the steps of;

generating a first value by counting pixels corresponding to the edges;

generating a second value by counting pixels corresponding to the spaces;

comparing the first value and the second value with predetermined first threshold and second threshold generated by summing up a threshold of a previous row, and an average number of pixels verified to edges and spaces in the previous row, respectively, to verify the counted pixels according to the results of the comparison.

11. The method of claim 10, wherein the step of comparing the first value and the second value is verified when the first value is greater than the first threshold and the second value is less than the second threshold.

12. The method of claim 10, wherein the step of discriminating between the spaces and the discontinuous edges detected includes verifying the counted pixels to the spaces when the first value is less than the first threshold and the second value is greater than the second threshold.

13. The method of claim 10 wherein the step of eliminating the noise includes replacing values of the noise with values of the spaces of the image.

14. The method of claim 10, wherein the step of connecting the discontinuous edges includes replacing values of the discontinuous edges with values of the edges of the image.