Image processing apparatus and method

• US 6,556,711 B2
• Filed: 09/30/1998
• Issued: 04/29/2003
• Est. Priority Date: 12/28/1994
• Status: Expired due to Fees

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Ishida, Yoshihiro, Yoshizaki, Osamu, Koga, Shinichiro
• Primary Examiner(s): Wu, Jingge
• Application Number:
  US09/162,727
• Publication Number:
  US 20020064307A1
• Time in Patent Office: 1,672 Days
• Field of Search: 382/164, 382/173, 382/176, 382/177, 382/180, 382/298, 382/299, 358/453, 358/462, 358/464
• US Class Current: 382/173
• CPC Class Codes: G06T 2207/10008 from scanner, fax or copier
• CPC Class Codes: G06T 2207/30176 Document
• CPC Class Codes: G06T 3/40 Scaling of whole images or parts thereof, e.g. expanding or contracting
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/194 involving foreground-background segmentation
• CPC Class Codes: G06V 10/56 relating to colour
• CPC Class Codes: G06V 30/414 Extracting the geometrical structure, e.g. layout tree; Block segmentation, e.g. bounding boxes for graphics or text

Abstract

A digital color image to be processed is inputted at a color image input step, image segments are extracted from the input color image at an image-segment extraction step, and the data of the extracted image segments is created. Next, image-segment components in each extracted image segment are discriminated at an image-segment discrimination step, and each extracted image segment is subjected to zoom processing, which conforms to the discriminated image-segment component, at an adaptive zoom step, thereby creating a zoomed image of the input image. The zoomed image that has been created is displayed, outputted as a hard copy or delivered to a transmission line at a color image output step. Accordingly, a color image in which image segments having different characteristics are mixed can be subjected to excellent processing.

Claims

1. An image processing method of extracting an image area from an image in which image areas having different components are mixed, comprising the steps of:

generating a reduced image from an inputted image;

extracting an image area from the reduced image;

calculating a position of the image area, extracted in said first extracting step, on the inputted image;

extracting an image area from the inputted image based on the calculated position;

detecting an image area, which is omitted in said generating step, based on the calculated position and the image area extracted in said second extracting step; and

unifying the image area extracted in said second extraction step and the detected image area.

2. The method according to claim 1, wherein said first extracting step comprises the steps of:

determining whether an image area to undergo processing in the input image has an image area which will be background;

separating the determined background image area; and

separating at least one image area from image areas other than the background image area,

wherein said first extracting step extracts image areas by recursively repeating said determining step and said separating steps in the image area obtained in said unifying step.

3. The method according to claim 1,

wherein said calculating step calculates the position from position information of the image area extracted in said first extracting step and a reduction ratio of the reduced image, and

wherein said second extracting step extracts the image area based upon a rectangular image area at the calculated position.

4. The method according to claim 1, wherein said second extracting step extracts the image area by referring to pixels of the inputted image that correspond to the image area extracted in said first extracting step.

5. The method according to claim 1, wherein said second extracting step comprises the steps of:

applying expansion processing to a shape of the image area extracted in said first extracting step; and

extracting the image area by referring to pixels of the inputted image that correspond to the expanded image area.

6. The method according to claim 1, wherein said second extracting step comprises the steps of:

binarizing the inputted image to a background and portions other than the background using a value of a background color of the image area extracted in said first extracting step; and

extracting the image area from the binarized image.

7. The method according to claim 1, wherein said generating step selects one pixel from each block comprising N×

N (where N is a natural number) pixels obtained by partitioning the inputted image, and representing this block by the one pixel, thereby creating a reduced image having a reduction ration of 1/N.

8. The method according to claim 1, wherein said generating step obtains an average value of pixels of each block comprising N×

N (where N is a natural number) pixels obtained by partitioning the inputted image, and representing this block by the average value, thereby creating a reduced image having a reduction ration of 1/N.

9. An image processing apparatus for extracting an image area from an image in which image areas having different components are mixed, comprising:

a reduction unit adapted to generate a reduced image from an inputted image;

a first extractor adapted to extract an image area from the reduced image;

a calculator adapted to calculate a position of the image area, extracted by said first extractor, on the inputted image;

a second extractor adapted to extract an image area from the inputted image based on the calculated position;

a detector adapted to detect an image area, which is omitted by said reduction unit, based on the calculated position and the image area extracted by said second extractor; and

a unification unit adapted to unify the image area extracted by said second extractor and the detected image area.

10. A computer program product storing a computer-readable medium comprising computer program code, which executes an image processing method for extracting an image area from an image in which image areas having different components are mixed, wherein the method comprises the steps of:

generating a reduced image from an inputted image;

extracting an image area from the reduced image;

calculating a position of the image area, extracted in said first extracting step, on the inputted image;

extracting an image area from the inputted image based on the calculated position;

detecting an image area, which is omitted in said generating step, based on the calculated position and the image area extracted in said second extracting step; and

unifying the image area extracted in said second extracting step and the detected image area.