Candidate article recognition with assignation of reference points and respective relative weights

• US 5,093,867 A
• Filed: 08/20/1990
• Issued: 03/03/1992
• Est. Priority Date: 07/22/1987
• Status: Expired due to Term

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Hori, Kazuhito, Nagata, Yasuyuki
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): CAMMARATA, MICHAEL
• Application Number:
  US07/569,513
• Time in Patent Office: 561 Days
• Field of Search: 382/8, 382/23, 382/30, 382/34, 382/14, 382/45, 382/50, 382/65, 382/36
• US Class Current: 382/141
• CPC Class Codes: G06T 2207/30164 Workpiece; Machine component
• CPC Class Codes: G06T 7/001 using an image reference approach
• CPC Class Codes: G06V 10/507 Summing image-intensity values; Histogram projection analysis
• CPC Class Codes: G06V 2201/02 Recognising information on displays, dials, clocks

Abstract

A high-speed image recognition system operates in response to a binary digitized picture image signal. The system comprises a first memory for storing a bi-level reference image pattern therein, a selector coupled to the first memory for manually designating a plurality of selected points of the bi-level reference image pattern, a detector coupled to the selector for detecting signal levels at each of the points selected by the selector, a second memory coupled to the detector for storing therein detected signal levels obtained from the detector, and a comparator coupled to the second memory for comparing the detected signal levels at the selected points of the bi-level reference image pattern with signal levels at the same points of a candidate image pattern, whereby an object corresponding to the candidate image pattern is recognized.

Claims

1. Image recognition apparatus for recognizing a condition of a candidate article with respect to a reference condition, said apparatus comprising:

first memory means for storing a bi-level reference image;

display means connected to said first memory means for displaying said reference image;

selecting means connected to said first memory means and said display means for manually designating a plurality of selected reference points of said reference image;

detecting means that has at least one adjustable parameter and is connected to said selecting means for detecting signal levels at each of said reference points;

second memory means connected to said detecting means for storing said detected signal levels; and

comparison means connected to said second memory means for making a comparison of said detected signal levels with bi-level signals at points that correspond to said reference points but are proper to a bi-level candidate image corresponding to a candidate article;

said comparison indicating a condition of said candidate article with respect to a corresponding condition of said reference image, and said display means displaying the results of said comparison;

wherein each of said points correspond to a plurality of picture elements and said detecting means comprises means for adjustably assigning relative weights to said picture elements.

2. Image recognition apparatus according to claim 1, further comprising:

scanning means for scanning a candidate article and a reference article and generating image signals corresponding thereto; and

bi-level converting means coupled to said scanning means for converting said image signals into said bi-level signals.

3. Image recognition apparatus according to claim 1, wherein said selecting means includes means for manually defining an absolute position of said reference image and an associated first set of selected reference points, for indicating an absolute position of said candidate image, and for automatically defining an associated second set of selected reference points, the reference points of said second set respectively corresponding to the reference points of the said first set, whereby the comparison made by said comparison means is indicative of the identity or non-identity of said candidate image with respect to said reference image.

4. Image recognition apparatus according to claim 1, wherein said selecting means includes calculating means for defining a center of gravity of said candidate article, and said points corresponding to said reference points are selected with respect to said center of gravity.

5. Image recognition apparatus according to claim 1, wherein said selecting means includes calculating means for defining a center of gravity and principal axis of inertia of said candidate article, and said points corresponding to said reference points are selected with respect to said center of gravity and principal axis of inertia.