Method for recognizing concavities in an image subject to character recognition

• US 5,159,645 A
• Filed: 10/29/1991
• Issued: 10/27/1992
• Est. Priority Date: 05/12/1988
• Status: Expired due to Term

• Current Assignee: YOZAN, Inc.
• Original Assignee: Ezel, Inc.
• Inventors: Kumagai, Ryohei
• Primary Examiner(s): Couso, Jose L.
• Application Number:
  US07/784,126
• Time in Patent Office: 364 Days
• Field of Search: 382/21, 382/22, 382/25, 382/26, 382/48
• US Class Current: 382/198
• 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

Abstract

An image processing method for finding concavity of a configuration comprising steps; arranging boundary pixels of the configuration according to chain order with indicating top points on convex hull of the configuration; and determining that a concavity is defined by the boundary pixels when a boundary pixel of a chain code is found, which chain code indicates a concavity.

Claims

1. An image processing method for recognizing a configuration in an image from a characteristics representation of the configuration comprising the steps of:

representing the image as digital pixels;

generating a convex hull for the configuration, said convex hull delineated by a series of pixels connecting top points of the configuration;

generating a representation of the configuration as characteristics of the difference between the convex hull and the configuration, said characteristics including concavities in the configuration; and

wherein the step of generating a representation of the configuration includes steps of;

representing a contour of the configuration between two adjacent top points as a series of chain codes, said top points defining a straight line segment;

calculating separation distances between the contour of the configuration and the straight line segment; and

generating a representation of the contour as a concavity when a separation distance is greater than a predetermined value.

2. An image processing method according to claim 1, wherein boundary pixels of said configuration are divided into groups of possible chain codes included in a area other than a concavity.

3. An image processing method according to claim 2, wherein said groups comprise four quadrants.

4. An image processing method according to claim 3, wherein said quadrants are defined by most upper, most leftward, most lower and most rightward boundary pixels.

5. An image processing method according to claim 4, wherein said possible chain codes are defined as follows;

between most upper and most leftward boundary pixels;

chain code=1, 5 or 6

between most left and most lower boundary pixels;

chain code=6, 7 or 0

between most lower and most rightward boundary pixels;

chain code=0, 1 or 2

between most rightward and most upper boundary pixels;

chain code=2, 3 or 4.

6. An image processing method according to claim 2, wherein the step of calculating separation distances includes steps of:

calculating distance increments for each chain code; and

integrating distance increments of pixels as processed from one of said top points toward the other.

7. An image processing method according to claim 2, wherein number of boundary pixels are counted between said two top points on evaluating chain codes or on calculating distances, and total number of boundary pixels are counted when said boundary pixels are found to define a concavity, so as to calculate inner peripheral length of a concavity.