Method of and apparatus for detecting objects in an assembly line

• US 4,995,091 A
• Filed: 10/25/1989
• Issued: 02/19/1991
• Est. Priority Date: 10/25/1989
• Status: Expired due to Fees

• Current Assignee: Ford Global Technologies, LLC (Ford Motor Company)
• Original Assignee: Mazda Motor Manufacturing
• Inventors: Shimbara, Yoshimi
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): Fox, David
• Application Number:
  US07/426,231
• Time in Patent Office: 482 Days
• Field of Search: 382/50, 382/51, 382/52, 382/37
• US Class Current: 382/272
• CPC Class Codes: B07C 5/3422 using video scanning devices, e.g. TV-cameras
• CPC Class Codes: G06V 10/28 Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns

Abstract

A method and apparatus for identifying marking holes in an object moving along a conveyor or the like, in which television image signals of the identifying marking holes are classified into signals above and below a threshold level voltage and then made into a monochrome image of the identifying marking holes which is then compared with a pre-memorized image of the marking hole. The invention involves a particular way of selecting the threshold level voltages by using a threshold level voltage from among threshold level voltages used to identify the marking holes in preceding objects according to the frequency of occurrence of the threshold level voltages, and for the threshold level voltage of the second identifying marking hole, initially calculating a target threshold level voltage, and then modifying it if necessary by reference to threshold level voltages used to identify the second marking hole in preceding objects, so as to make the identification of the second hole quicker.

Claims

1. A method of detecting the presence of an object moving along an assembly line among a succession of objects and which objects have at least two identifying features thereon, comprising:

obtaining an image of the first identifying feature of one of the objects constituted by a plurality of voltage signals of more than two different voltages;

choosing a first threshold level voltage for the first identifying feature of said one object;

classifying the signals of said plurality of voltage signals for said first identifying feature of said one object into signals greater than the first threshold level voltage and signals less than the first threshold level voltage;

forming from said thus classified signals a monochrome image;

comparing said monochrome image with a pre-memorized image of the first identifying feature constituted by a plurality of voltage signals for determining whether the monochrome image corresponds to the pre-memorized image;

when said image of said first identifying feature corresponds with the pre-memorized image, designating the threshold level voltage at which the correspondence occurs as matching threshold level voltage A_n ;

obtaining an image of the second identifying feature of said one object constituted by a plurality of voltage signals of more than two different voltages;

calculating a target threshold level voltage B_nt for the second identifying feature of said one object according the expression;

space="preserve" listing-type="equation">B.sub.nt =f(A.sub.n *, B.sub.n-1 *, A.sub.n)

where A_n * is the average matching threshold level voltage for the first identifying features of preceding objects in the succession and B_n-1 * is the average matching threshold level voltage for the second identifying features of the preceding objects in the succession;

classifying the signals of said plurality of voltage signals for said second identifying feature into signals greater than the target threshold level voltage and signals less than the target threshold level voltage;

forming from said thus classified signals a monochrome image of second identifying feature;

comparing said monochrome image of the second identifying feature with a pre-memorized image of the second identifying feature constituted by a plurality of voltage signals for determining whether the monochrome image of the second identifying feature corresponds to the pre-memorized image of the second identifying feature;

when said image of said second identifying feature corresponds with the pre-memorized image, designating the threshold level voltage at which correspondence occurs as matching threshold level voltage B_n ;

obtaining an image of the first identifying feature of the next succeeding object and carrying out the steps of choosing a first threshold level voltage, classifying the signals, forming the monochrome image, and comparing the image for the last-mentioned first identifying feature to obtain a matching threshold level voltage A_n+1 for said last mentioned first identifying feature,

calculating a target threshold level voltage B_n+1)T for the second identifying feature on said next succeeding object according to the expression;

space="preserve" listing-type="equation">B.sub.n+1)T =f(A.sub.n+1 *, B.sub.n *, A.sub.n+1)

where A_n+1 * is the average matching threshold level voltage for the first identifying features of the succession of objects including matching threshold level voltage A_n+1, B_n * is the average matching threshold level voltage for the second identifying features of the succession of objects including matching threshold level voltage B_n ; and

obtaining an image of the second identifying feature of the next succeeding object and carrying out the steps of classifying the signals using the newly calculated target threshold level voltage, forming the monochrome image and comparing the image for the last mentioned second identifying feature.

2. A method as claimed in claim 1 in which said steps of choosing a first threshold level voltage, classifying and comparing comprise;

initially choosing as a first threshold level voltage a voltage from among a voltage based on past experience and a voltage determined experimentally; and

when the determination in said comparison step of comparing the monochrome image of the first identifying feature with the pre-memorized image of the first identifying feature shows that the monochrome image of the first identifying feature does not correspond to the pre-memorized image of the first identifying feature, choosing a further first threshold level voltage corresponding to a matching threshold level voltage for first identifying features for preceding objects in the succession of objects;

classifying the signals of said plurality of voltage signals for the first identifying feature into signals greater than the newly chosen further first threshold level voltage and signals less than the newly chosen further first threshold level voltage;

forming from said thus classified signals a monochrome image of the first identifying feature;

comparing the thus obtained image with said pre-memorized image of the first identifying feature for determining whether the monochrome image obtained with the further first threshold level voltage corresponds to the pre-memorized image of the first identifying feature; and

when the last mentioned comparison shows that the monochrome image obtained with the further first threshold level voltage does not correspond to the pre-memorized image, repeating the further first threshold level voltage choosing step, the classifying step, the image forming step and the comparison step until the monochrome image obtained corresponds with the pre-memorized image of the first identifying feature.

3. A method as claimed in claim 2 in which the step of choosing the further first threshold level voltage comprises choosing the threshold level voltage of the matching threshold voltages which occur with the greatest frequency among said matching threshold voltage levels for the first identifying feature for preceding objects, and the step of repeating the further first threshold level voltage choosing step comprises choosing the threshold level voltage of the matching threshold voltages which occur with the next greatest frequency among said matching threshold level voltages for the first identifying feature for the preceding objects.

4. A method as claimed in claim 2 further comprising:

when the determination in said comparison step of comparing the monochrome image of the second identifying feature with the pre-memorized image of the second identifying feature shows that the monochrome image of the second identifying feature does not correspond to the pre-memorized image of the second identifying feature, choosing a second threshold level voltage corresponding to a matching threshold level voltage for second identifying features of preceding objects in the succession of objects and which voltage is the next adjacent said target threshold level voltage;

classifying the signals of said plurality of voltage signals into signals greater than the newly chosen second threshold level voltage and signals less than the newly chosen second threshold level voltage;

forming from said thus classified signals a monochrome image of the second identifying feature;

comparing the thus obtained image of the second identifying feature with said pre-memorized image of the second identifying feature for determining whether the monochrome image of the second identifying feature corresponds to the pre-memorized image of the second identifying feature; and

when the last mentioned comparison shows that the monochrome image of the second identifying feature does not correspond to the pre-memorized image of the second identifying feature, choosing a further second threshold level voltage and repeating the classifying step, the image forming step and the comparison step until the monochrome image of the second identifying feature corresponds with the pre-memorized image of the second identifying feature.

5. A method as claimed in claim 4 in which the next adjacent threshold level voltage is the next higher adjacent threshold level voltage.

6. A method as claimed in claim 4 in which the next adjacent threshold level voltage is the next lower adjacent threshold level voltage.

7. A method as claimed in claim 4 in which the step of choosing the further second threshold level voltage comprises choosing the threshold level voltage among the matching threshold level voltages which occurs with the next greatest frequency to the matching threshold level voltage chosen for said second threshold level voltage.

8. An apparatus for detecting the presence of an object moving along an assembly line among a succession of objects and which objects have at least two identifying features thereon, comprising:

means for obtaining an image of the first identifying feature of one of the objects constituted by a plurality of voltage signals of more than two different voltages;

means for choosing a first threshold level voltage for the first identifying feature of said one object;

means for classifying the signals of said plurality of voltage signals for said first identifying feature of said one object into signals greater than the first threshold level voltage and signals less than the first threshold level voltage;

means for forming from said thus classified signals a monochrome image;

means for comparing said monochrome image with a pre-memorized image of the first identifying feature constituted by a plurality of voltage signals for determining whether the monochrome image corresponds to the pre-memorized image, and when said image of said first identifying feature corresponds with the pre-memorized image, designating the threshold level voltage at which the correspondence occurs as matching threshold level voltage A_n ;

means for obtaining an image of the second identifying feature of said one object constituted by a plurality of voltage signals of more than two different voltages;

means for calculating a target threshold level voltage B_nt for the second identifying feature of said one object according to the expression;

space="preserve" listing-type="equation">B.sub.nt =f(A.sub.n *, B.sub.n-1 *, A.sub.n)

where A_n * is the average matching threshold level voltage for the first identifying features of preceding objects in the succession and B_n-1 * is the average matching threshold level voltage for the second identifying features of the preceding objects in the succession;

means for classifying the signals of said plurality of voltage signals for said second identifying feature into signals greater than the target threshold level voltage and signals less than the target threshold level voltage;

means for forming from said thus classified signals a monochrome image of second identifying feature;

means for comparing said monochrome image of the second identifying feature with a pre-memorized image of the second identifying feature constituted by a plurality of voltage signals for determining whether the monochrome image of the second identifying feature corresponds to the pre-memorized image of the second identifying feature, and when said image of said second identifying feature corresponds with the pre-memorized image, designating the threshold level voltage at which correspondence occurs as matching threshold level voltage B_n ;

means obtaining an image of the first identifying feature of the next succeeding object and for carrying out the steps of choosing a first threshold level voltage, classifying the signals, forming the monochrome image, and comparing the image for the last mentioned first identifying feature to obtain a matching threshold level voltage A_n+1 for said last mentioned first identifying feature,

means for calculating a target threshold level voltage B.sub.(n-1)T for the second identifying feature on said next succeeding object according to the expression;

space="preserve" listing-type="equation">B.sub.n+1)T =f(A.sub.n+1 *, B.sub.n *, A.sub.n+1)

where A_n+1 * is the average matching threshold level voltage for the first identifying features of the succession of objects including matching threshold level voltage A_n+1, B_n * is the average matching threshold level voltage for the second identifying features of the succession of objects including matching threshold level voltage B_n ; and

means for obtaining an image of the second identifying feature of the next succeeding object and carrying out the steps of classifying the signals using the newly calculated target threshold level voltage, forming the monochrome image and comparing the image for the last mentioned second identifying feature.