Object inspection method employing selection of discerning features using mahalanobis distances

• US 5,392,364 A
• Filed: 05/20/1992
• Issued: 02/21/1995
• Est. Priority Date: 05/23/1991
• Status: Expired due to Term

• Current Assignee: Matsushita Electric Industrial Company Limited (Panasonic Holdings Corporation)
• Original Assignee: Matsushita Electric Industrial Company Limited (Panasonic Holdings Corporation)
• Inventors: Nakao, Masaya, Yokoyama, Haruhiko
• Primary Examiner(s): Razavi, Michael T.
• Assistant Examiner(s): Fox, David
• Application Number:
  US07/885,837
• Time in Patent Office: 1,007 Days
• Field of Search: 382/16, 382/14, 382/15, 382/30, 382/34, 382/36, 382/37, 382/39
• US Class Current: 382/190
• CPC Class Codes: G06V 10/75 Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries

Abstract

A method for discerning whether an object to be inspected is acceptable or not is based on feature values with respect to a binary-coded image of the object. The method includes the steps of coding image data of the object into binary digits to obtain the binary-coded image, calculating at least three feature values based on a predetermined sample group of acceptable objects and a predetermined sample group of unacceptable objects, obtaining a Mahalanobis'"'"' generalized distance between the sample groups of the acceptable objects and the unacceptable objects with respect to each of the calculated feature values, comparing each of the distances with a first predetermined value and then selecting as a first representative feature value the distance which is not smaller than the first predetermined value, obtaining a Mahalanobis'"'"' generalized distance between groups of acceptable objects and unacceptable objects with respect to the feature values except for the feature value selected as the first representative feature value and the first representative feature value, and comparing each of the distances with a second predetermined value and then selecting as a second representative feature value the distance which is not smaller than the second predetermined value, so that it is discerned whether the object is acceptable or not based on the first and/or first and second feature values with respect to the binary-coded image of the object.

Claims

1. In an image inspection process in which an object to be recognized is viewed with an image pick-up device, an output of the image pick-up device is converted into binary coded image data, a plurality of shape feature values of the binary coded image data are calculated, and an acceptability of the object is automatically discerned by comparing the plurality of shape feature values with predetermined values, the improvement comprising a preprocessing method for prioritizing the plurality of shape feature values used in the image inspection process so as to minimize a number of the shape feature values which must be processed to discern the acceptability of the object, said preprocessing method comprising:

viewing a predetermined group of sample acceptable objects and a predetermined group of sample unacceptable objects with an image pick-up device to obtain respective image outputs of the sample acceptable objects and the sample unacceptable objects;

converting the image outputs of the image pick-up device into a binary coded image data for each of the sample acceptable objects and sample unacceptable objects;

computing at least three shape feature values of each of the binary coded image data for each of the sample acceptable objects and sample unacceptable objects;

computing a first Mahalanobis'"'"' generalized distance between the binary coded image data of the sample acceptable objects and the binary coded image data of the sample unacceptable objects with respect to each of the at least three shape feature values;

designating as a first representative shape feature value a first one of the at least three shape feature values having a first Mahalanobis'"'"' generalized distance which is not smaller than a first value;

computing a second Mahalanobis'"'"' generalized distance between the binary coded image data of the sample acceptable objects and the binary image data of the sample unacceptable objects with respect to each of the at least three shape feature values, except for the first one of the at least three shape feature values, relative to the first one of the at least three shape feature values; and

designating as a second representative shape feature value a second one of the at least three shape feature values, except for the first one of the at least three shape feature values, having a second Mahalanobis'"'"' generalized distance which is not smaller than a second value;

wherein the first and second representative shape feature values are given priority in the image inspection process to discern the acceptability of an object to be inspected.

2. In an image inspection process as claimed in claim 1, wherein the first representative shape feature value is designated as a first one of the at least three shape feature values having a maximum first Mahalanobis'"'"' distance.

3. In an image inspection process as claimed in claim 2, wherein the second representative shape feature value is designated as a second one of the at least three shape feature values, except for the first one of the at least three shape feature values, having a maximum second Mahalanobis'"'"' distance.

4. In an image inspection process as claimed in claim 1, wherein the second representative shape feature value is designated as a second one of the at least three shape feature values, except for the first one of the at least three shape feature values, having a maximum second Mahalanobis'"'"' distance.

5. An image inspection method comprising a preprocess for prioritizing a plurality of shape feature values and an inspection process for discerning the acceptability of an object using the plurality of shape feature values prioritized in said preprocess;

said preprocess comprising,

(a) viewing a predetermined group of sample acceptable objects and a predetermined group of sample unacceptable objects with an image pick-up device to obtain respective image outputs of the sample acceptable objects and the sample unacceptable objects,

(b) converting the image outputs of the image pick-up device into a binary coded image data for each of the sample acceptable objects and sample unacceptable objects,

(c) computing at least three shape feature values of each of the binary coded image data for each of the sample acceptable objects and sample unacceptable objects,

(d) computing a first Mahalanobis'"'"' generalized distance between the binary coded image data of the sample acceptable objects and the binary coded image data of the sample unacceptable objects with respect to each of the at least three shape feature values,

(e) designating as a first representative shape feature value a first one of the at least three shape feature values having a first Mahalanobis'"'"' generalized distance which is not smaller than a first value,

(f) computing a second Mahalanobis'"'"' generalized distance between the binary coded image data of the sample acceptable objects and the binary image data of the sample unacceptable objects with respect to each of the at least three shape feature values, except for the first one of the at least three shape feature values, relative to the first one of the at least three shape feature values, and

(g) designating as a second representative shape feature value a second one of the at least three shape feature values, except for the first one of the at least three shape feature values, having a second Mahalanobis'"'"' generalized distance which is not smaller than a second value;

said inspection process comprising,

(a) viewing the object to be recognized with an image pick-up device to obtain an image output of the object to be inspected,

(b) converting the image output of the image pick-up device into binary coded image data of the object to be inspected,

(c) computing the first and second representative shape feature values of the binary coded image data of the object to be inspected, and

(d) discerning an acceptability of the object to be inspected by comparing at least the computed first representative shape feature value, or the computed first and second representative shape feature values, of the binary coded image data with predetermined values.

6. An image inspection method as claimed in claim 5, wherein the first representative shape feature value is designated as a first one of the at least three shape feature values having a maximum first Mahalanobis'"'"' distance.

7. An image inspection method as claimed in claim 6, wherein the second representative shape feature value is designated as a second one of the at least three shape feature values, except for the first one of the at least three shape feature values, having a maximum second Mahalanobis'"'"' distance.

8. An image inspection method as claimed in claim 5, wherein the second representative shape feature value is designated as a second one of the at least three shape feature values, except for the first one of the at least three shape feature values, having a maximum second Mahalanobis'"'"' distance.