IMAGE PROCESSING APPARATUS, IMAGE PROCESSING METHOD, AND COMPUTER-READABLE, NON-TRANSITORY MEDIUM

• US 20140212049A1
• Filed: 12/03/2013
• Published: 07/31/2014
• Est. Priority Date: 01/30/2013
• Status: Active Grant

• Current Assignee: PFU Limited (Fujitsu Limited)
• Original Assignee: PFU Limited (Fujitsu Limited)
• Inventors: Honda, Keiji
• Granted Patent: US 9,117,283 B2
• Time in Patent Office: Days
• US Class Current: 382/199
• CPC Class Codes: G06T 2207/30176 Document
• CPC Class Codes: G06T 7/13 Edge detection
• CPC Class Codes: G06T 7/181 involving edge growing; involving edge linking
• CPC Class Codes: G06V 10/46 Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features colour feature extraction G06V10/56
• CPC Class Codes: G06V 10/48 by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation

Abstract

Provided are an image processing apparatus, image processing method and a computer-readable, non-transitory medium that can accurately detect document edges from a readout image. The image processing apparatus includes a first edge pixel detector for detecting a plurality of first edge pixels from an input image, a straight line detector for detecting a straight line from the first edge pixels, a classifying module for classifying the first edge pixels into on-line edge pixels and non-on-line edge pixels, a second edge pixel detector for detecting second edge pixels located between two of the on-line edge pixels, when there is any one of the non-on-line edge pixels between two of the on-line edge pixels, and a determining module for determining whether the non-on-line edge pixel represents a document edge, based on whether the two on-line edge pixels are connected by the second edge pixels via the non-on-line edge pixel.

Claims

1. An image processing apparatus comprising:

a first edge pixel detector for detecting a plurality of first edge pixels from an input image;

a straight line detector for detecting a straight line from the plurality of first edge pixels;

a classifying module for classifying the plurality of first edge pixels into on-line edge pixels which are located within a predetermined distance from the straight line and non-on-line edge pixels which are not located within the predetermined distance;

a second edge pixel detector for detecting second edge pixels located between two of the on-line edge pixels, when there is any one of the non-on-line edge pixels between two of the on-line edge pixels; and

a determining module for determining whether the non-on-line edge pixel represents a document edge or not, based on whether or not the two on-line edge pixels are connected by the second edge pixels via the non-on-line edge pixel.

2. The image processing apparatus according to claim 1, wherein the second edge pixel detector repeatedly performs processing for detecting one second edge pixel by performing a search within a predetermined range from one of the two on-line edge pixels and for detecting the next second edge pixel by further performing a search within a predetermined range from the detected one second edge pixel, and

wherein the determining module determines that the two on-line edge pixels are connected by the second edge pixels via the non-on-line edge pixel, when the other of the two on-line edge pixels is located within a predetermined range of any one of the second edge pixels detected within a predetermined number of times.

3. The image processing apparatus according to claim 2, wherein the second edge pixel detector calculates a pixel value difference between each pixel detected within the predetermined range and each pixel located adjacent to the detected pixel, and detects a pixel whose pixel value difference is the largest within the predetermined range as the second edge pixel.

4. The image processing apparatus according to claim 3, wherein the second edge pixel detector gives weight to the pixel value difference according to a direction in which the second edge pixel was detected, for all pixels located within a predetermined range from the detected second edge pixel, and detects a pixel whose weighted pixel value difference is the largest as the second edge pixel.

5. The image processing apparatus according to claim 1, wherein the determining module determines that the second edge pixel also represents a document edge when it is determined that the non-on-line edge pixel represents a document edge.

6. The image processing apparatus according to claim 1, wherein the determining module determines that pixels lying on a straight line joining the two on-line edge pixels represent a document edge when two of the on-line edge pixels detected for one straight line detected by the straight line detector are unable to be connected by the second edge pixels via the non-on-line edge pixel.

7. The image processing apparatus according to claim 1, wherein the determining module determines that pixels lying on a straight line joining an intersection of the two straight lines to each one of two of the on-line edge pixels represent a document edge when the two on-line edge pixels respectively detected for two straight lines detected by the straight line detector are unable to be connected by the second edge pixels via the non-on-line edge pixel.

8. The image processing apparatus according to claim 1, further comprising a correction module for correcting pixels located between the straight line detected by the straight line detector and the non-on-line edge pixel and the second edge pixels determined to represent the document edge, by using the non-on-line edge pixel and the second edge pixels or using pixels located on a side of the non-on-line edge pixel and the second edge pixels opposite from the straight line.

9. The image processing apparatus according to claim 8, wherein the correction module corrects pixels located between the straight line detected by the straight line detector and the on-line edge pixels, by using the on-line edge pixels or using pixels located on a side of the on-line edge pixels opposite from the straight line.

10. An image processing method comprising:

detecting a plurality of first edge pixels from an input image;

detecting a straight line from the plurality of first edge pixels;

classifying the plurality of first edge pixels into on-line edge pixels which are located within a predetermined distance from the straight line and non-on-line edge pixels which are not located within the predetermined distance;

detecting second edge pixels located between two of the on-line edge pixels when there is any one of the non-on-line edge pixels between two of the on-line edge pixels; and

determining, by a computer, whether the non-on-line edge pixel represents a document edge or not, based on whether or not the two on-line edge pixels are connected by the second edge pixels via the non-on-line edge pixel.

11. A computer-readable, non-transitory medium storing a computer program, wherein the computer program causes a computer to execute a process, the process comprising:

detecting a plurality of first edge pixels from an input image;

detecting a straight line from the plurality of first edge pixels;

classifying the plurality of first edge pixels into on-line edge pixels which are located within a predetermined distance from the straight line and non-on-line edge pixels which are not located within the predetermined distance;

detecting second edge pixels located between two of the on-line edge pixels when there is any one of the non-on-line edge pixels between two of the on-line edge pixels; and

determining whether the non-on-line edge pixel represents a document edge or not, based on whether or not the two on-line edge pixels are connected by the second edge pixels via the non-on-line edge pixel.