IMAGE PROCESSING METHOD AND APPARATUS

• US 20030076539A1
• Filed: 05/04/1999
• Published: 04/24/2003
• Est. Priority Date: 05/11/1998
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: NAKAJIMA, SATORU
• Granted Patent: US 6,690,492 B2
• Time in Patent Office: Days
• US Class Current: 358/2.1
• CPC Class Codes: G06T 2207/10008 from scanner, fax or copier
• CPC Class Codes: G06T 2207/30176 Document
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06V 30/413 Classification of content, e.g. text, photographs or tables
• CPC Class Codes: H04N 1/3875 combined with enlarging or reducing enlarging or reducing per se H04N1/393

Abstract

According to this invention, in order to efficiently output images, when an image and its attribute are selected, and an extraction output instruction of an image region with the selected attribute is issued, the selected image is segmented into image regions with different attributes, and an image region with the selected attribute is extracted and displayed from the segmented image regions.

Claims

1. An image processing method comprising:

a selection instruction step of instructing to select an image and image attribute, and to extract and output an image region with a selected attribute; and

an extraction output step of segmenting an image selected in the selection instruction step into image regions with different attributes, and extracting and outputting an image region with the attribute selected in the selection instruction step from the segmented image regions.

2. The method according to claim 1, wherein the extraction output step includes:

a step of recognizing a continuous pixel group having a predetermined density in an image as a rectangular region;

a step of identifying an attribute of the rectangular region using a width, height, area, and pixel density of the rectangular region; and

a step of determining the attribute of the rectangular region by computing and comparing local histograms of the rectangular region.

3. An image processing method comprising:

an image selection step of selecting an image;

a text extraction instruction step of instructing to extract an image region with a text attribute; and

a text output step of segmenting the image selected in the image selection step into image regions having different attributes, and extracting and outputting an image region with the text attribute from the segmented image regions in accordance with an instruction issued in the text extraction instruction step.

4. The method according to claim 3, wherein the text output step comprises a text display step for displaying the extracted image region.

5. The method according to claim 4, further comprising:

a displayed text change instruction step of instructing to change the image region displayed in the text display step to another image region with a text attribute; and

a display change step of changing the image region to be displayed in accordance with an instruction issued in the displayed text change instruction step.

6. The method according to claim 4, wherein the text display step is capable of displaying a plurality of image regions with a text attribute.

7. The method according to claim 6, further comprising:

a selection instruction step of selecting an image region which is included in a pre-segmentation image requested to display, from the image regions displayed in the text display step, and instructing to display the pre-segmentation image; and

a pre-segmentation image display step of retrieving a pre-segmentation image including the image region selected in the selection instruction step, and displaying the pre-segmentation image.

8. The method according to claim 3, wherein the text output step includes:

a step of recognizing a continuous pixel group having a predetermined density in an image as a rectangular region;

a step of identifying an attribute of the rectangular region using a width, height, area, and pixel density of the rectangular region; and

a step of determining the attribute of the rectangular region by computing and comparing local histograms of the rectangular region.

9. An image processing apparatus comprising:

image input means for inputting image information;

segmentation means for segmenting an image input by said image input means into image regions having different attributes;

extraction means for extracting an image region having a predetermined attribute from the image regions segmented by said segmentation means; and

output means for outputting an image of the image region extracted by said extraction means.

10. The apparatus according to claim 9, wherein said output means comprises display means.

11. The apparatus according to claim 9, wherein the predetermined attribute is a text attribute.

12. The apparatus according to claim 9, wherein said image input means comprises image scanning means for scanning image information.

13. The apparatus according to claim 9, wherein said output means can retrieve and output the input image from the extracted image region.

14. The apparatus according to claim 9, wherein said segmentation means includes:

recognition means for recognizing a rectangle by detecting continuous pixels having a predetermined density;

identification means for identifying an attribute of the rectangle using a width, height, area, and pixel density of an area defined by the rectangle; and

determination means for determining the attribute of the region by computing and comparing local histograms of the region.

15. A computer readable memory storing:

a program module for acquiring an image &

image attribute selection instruction, and an extraction output instruction of an image region having a selected attribute; and

a program module for segmenting an image based on the selection instruction acquired by the program into image regions having different attributes, and extracting and outputting an image region having an attribute based on the selection instruction acquired by the program from the segmented image regions.