Mark quality inspection apparatus and method

• US 5,859,923 A
• Filed: 01/09/1997
• Issued: 01/12/1999
• Est. Priority Date: 12/29/1992
• Status: Expired due to Term

• Current Assignee: Cognex Corporation
• Original Assignee: Cognex Corporation
• Inventors: Petry, John P. III, Picard, Len L.
• Primary Examiner(s): Stamber, Eric W.
• Application Number:
  US08/785,822
• Time in Patent Office: 733 Days
• Field of Search: 364/551.01, 364/552, 364/559, 364/525, 382/141, 382/145, 382/146, 382/147, 382/148, 382/149, 382/209, 382/218
• US Class Current: 382/141
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/30144 Printing quality
• CPC Class Codes: G06T 7/001 using an image reference approach
• CPC Class Codes: G06V 10/993 Evaluation of the quality of the acquired pattern

Abstract

A machine vision system for inspection of a mark (such as a multi-character mark) on an IC device as shown. The system provides search and defect analysis for the mark, individual characters of the mark, and the foreground and the background of the mark. The system provides for search and defect analysis reports thereof. It includes components for operator training search and defect analysis models for the whole mark, and for automatic training of such models for individual characters of the mark.

Claims

1. Apparatus for machine vision inspection of a candidate mark having at least one sub-component at a location within said candidate mark, said apparatus comprising:

video digitizing means for acquiring a candidate image of said candidate mark, and for acquiring at least one model image of an ideal mark;

search training means, responsive to said at least one model image, for generating a search model signal representative of said ideal mark as a whole, and of at least one sub-component of said ideal mark at a location within said ideal mark;

search means, responsive to said search model signal, and to said candidate image, for determining selected attributes of said candidate mark as a whole, and said at least one sub-component of said candidate mark, and for generating a signal representative of said selected attributes, said selected attributes including relative location and match score;

defect analysis training means, responsive to said at least one model image, for generating a defect signal representative of said ideal mark as a whole, and of said at least one sub-component of said ideal mark;

defect analysis means, responsive to said defect signal and to said candidate image, for generating a deviation signal indicative of a deviation of said candidate mark as a whole from said ideal mark as a whole, and indicative of a deviation of said at least one sub-component of said candidate mark from said at least one sub-component of said ideal mark;

partitioning means, cooperative with said search training means and said defect analysis training means, and responsive to said ideal mark, for generating signals representative of a plurality of respective locations of sub-components of said ideal mark,

said partitioning means including organizing means for organizing said plurality of sub-components into lines and sub-component positions within each of said lines corresponding to said plurality of respective locations within said candidate mark; and

output means, coupled to said search means and said defect analysis means, for generating a report signal representative of said selected attributes and said deviation of said candidate mark as a whole, and of said at least one sub-component of said candidate mark.

2. The apparatus of claim 1, wherein both said ideal mark and said candidate mark have a background and a foreground,

wherein said defect analysis training means includes background training means and foreground training means for generating a defect analysis model of at least a portion of said background and at least a portion of said foreground of said ideal mark,

wherein said defect analysis means includes background inspection means and foreground inspection means for inspecting at least a portion of said background and at least a portion of said foreground of said candidate mark,

and wherein said output means includes means for generating a report signal that includes a signal representative of the deviation relative to the ideal mark of at least a portion of said background and said foreground of said candidate mark.

3. The apparatus of claim 1, wherein said defect analysis means includes:

means for accommodating displacement of said at least one sub-component of said candidate mark with respect to said location of a corresponding sub-component of said ideal mark within said ideal mark.

4. A method for machine vision inspection of a candidate mark having at least one sub-component at a location within said candidate mark, said method comprising the steps of:

acquiring a candidate image of said candidate mark, and acquiring at least one model image of an ideal mark;

generating signals representative of a plurality of respective locations of a plurality of sub-components of said ideal mark, including the step of organizing said plurality of sub-components into lines and sub-component positions within each of said lines corresponding to said plurality of respective locations within said ideal mark;

generating a search model signal representative of said ideal mark as a whole, and of at least one sub-component of said plurality of sub-components of said ideal mark at a respective location within said ideal mark;

determining selected attributes of said candidate mark as a whole, and said at least one sub-component of said candidate mark, and generating a signal representative of said selected attributes, said selected attributes including relative location and match score;

generating a defect signal representative of said ideal mark as a whole, and of said at least one sub-component of said ideal mark;

generating a deviation signal indicative of a deviation of said candidate mark as a whole from said ideal mark as a whole, and indicative of a deviation of said at least one sub-component of said candidate mark from said at least one sub-component of said ideal mark; and

generating a report signal representative of said selected attributes and said deviation of said candidate mark as a whole, and of said at least one sub-component of said candidate mark.

5. The method of claim 4, wherein both said ideal mark and said candidate mark have a background and a foreground, further including the steps of:

generating a defect analysis model of at least a portion of said background and at least a portion of said foreground of said ideal mark;

inspecting at least a portion of said background and at least a portion of said foreground of said candidate marks; and

generating a report signal that includes a signal representative of the deviation relative to the ideal mark of at least a portion of said background and said foreground of said candidate mark.

6. The method of claim 4, wherein said step of generating a deviation signal includes the step of:

accommodating displacement of said at least one sub-component of said candidate mark with respect to said location of a corresponding sub-component of said ideal mark within said ideal mark.