Automatic package inspection method

• US 5,137,362 A
• Filed: 03/26/1990
• Issued: 08/11/1992
• Est. Priority Date: 03/26/1990
• Status: Expired due to Term

• Current Assignee: Freescale Semiconductor, Inc. (NXP Semiconductors NV)
• Original Assignee: Motorola, Inc. (Motorola Solutions, Inc.)
• Inventors: LeBeau, Christopher J.
• Primary Examiner(s): Rosenberger, Richard A.
• Assistant Examiner(s): Pham, Hoa
• Application Number:
  US07/498,776
• Time in Patent Office: 869 Days
• Field of Search: 356/237, 356/239, 356/128, 356/376, 356/388, 356/390, 356/394, 356/398, 356/391, 356/392, 356/393, 356/397, 356/2, 250/561, 250/563, 382/8, 382/9, 382/22, 382/34, 382/23, 382/54, 382/55, 358/101, 358/106, 358/107
• US Class Current: 382/145
• CPC Class Codes: G06T 2207/20192 Edge enhancement; Edge preservation
• CPC Class Codes: G06T 2207/30148 Semiconductor; IC; Wafer
• CPC Class Codes: G06T 7/0006 using a design-rule based approach

Abstract

A real time automatic visual semiconductor package inspection method is provided wherein a direction edge enhancement is performed on an image of the package. The direction edge enhanced image is dilated and correlated to a stored direction edge shape to identify all shapes of interest in a package under inspection. Also, anomalous shapes and uncorrelated direction edge shapes are identified and dilated. The dilated direction edge shape is analyzed using relatively simple mathematic techniques such as counting the number of shapes of a particular type, transforming shapes of interest to identify points of interest, and measuring relative position between the points of interest to determine acceptability of the semiconductor package. Also, size and location of anomalous shapes are calculated to determine acceptability of the package.

Claims

1. A method for automatically inspecting a semiconductor package, wherein the semiconductor package has a number of edge features, the method comprising:

obtaining an image of the package;

performing a direction edge enhancement to determine a plurality of package direction edges;

wherein the direction edge features in the image and identifying a direction for each edge feature;

building a list of point locations of the package direction edges; and

comparing the point locations against a predetermined list of acceptable point locations.

2. The method of claim 1 further comprising morphological dilation of the package direction edges before building the list of point locations.

3. The method of claim 1 further comprising:

locating leads which extend from the package by obtaining an image of the leads, determining a plurality of lead direction edges, building a list of point locations on the lead direction edges, and comparing the point locations of the leads to each other and to the point locations of the package direction edges to determine acceptable shape and alignment of the leads.

4. The method of claim 1 further comprising:

identifying anomalous shapes having direction edge points which are not included on the predetermined list;

calculating area and location of the anomalous shapes; and

determining acceptability of the semiconductor package from the area and location of the anomalous shapes.

5. A method for automatically inspecting a semiconductor package, wherein the semiconductor package has a plurality of edge features, the method comprising:

obtaining an image of a sample package and leads which extend from the sample package;

determining a direction edge shape for the sample package, wherein the step of determining the direction edge shape for the sample package further comprises identifying edge features in the image of the sample package and identifying a direction for each edge feature;

determining a direction edge shape for leads which extend from the sample package, wherein the step of determining the direction edge shape for the leads of the sample package further comprises identifying edge features in the image of the leads which extend from the sample package and identifying a direction for each edge feature;

storing the direction edge shape for the sample package and leads;

obtaining an image of a package under inspection;

determining a direction edge shape for the package under inspection, wherein the step of determining the direction edge shape for the package under inspection further comprises identifying edge features in the image of the package under inspection and identifying a direction for each edge feature;

comparing the direction edge shape of the package under inspection to the stored direction edge shape of the sample package;

determining a direction edge shape of the leads of the package under inspection, wherein the step of determining the direction edge shape for the leads of the package under inspection further comprises identifying edge features in the image of the leads which extend from the package under inspection and identifying a direction for each edge feature; and

comparing the direction edge shape of the leads of the package under inspection to the direction edge shape of the sample leads and to the direction edge shape of the package under inspection.

6. The method of claim 5 further comprising determining a direction edge point list for any direction edges of the package under inspection which do not match the direction edge shape of the sample package and performing an area and centroid analysis of the direction edge point list.

7. The method of claim 5 wherein after determining the direction edge shape of the package under inspection the shape of the package under inspection is mathematically transformed to compensate for rotation, illumination, and magnification errors.

8. The method of claim 5 wherein the steps of obtaining an image, determining a direction edge shape and comparing the direction edge shape for the package under inspection and for the leads of the package under inspection are performed repeatedly for a plurality of packages.

9. A method of inspecting a semiconductor package, wherein the semiconductor package has a number of edge features, the method comprising:

obtaining an image of the package;

performing direction edge enhancement to form a direction edge shape, wherein the step of performing a direction edge enhancement further comprises the steps of identifying edge features of the image of the package and identifying a direction for each edge feature;

dilating the direction edge shape; and

testing the dilated direction edge shape for correlation to a predetermined shape which has been previously stored.

10. The method of claim 9 further comprising identifying dilated direction edge shapes which correspond to locations of upper right, upper left, lower right, and lower left package edge shapes.

11. The method of claim 10 further comprising identifying dilated direction edge shapes which correspond to leads which extend from the package.

12. The method of claim 11 further comprising calculating the location of an x-axis which lies equidistant from the location of the upper and lower package edge shapes;

calculating the location of a y-axis which lies equidistant from the location of the left and right package edge shapes;

calculating a lead shape center location for each of the direction edge shapes which correspond to leads;

calculating a difference between each of the lead shape center locations and the x-axis;

calculating a difference between each of the lead shape center locations and the y-axis;

calculating a location of a best-fit line using the lead shape center locations;

subtracting the lead shape center locations from the best fit line to determine a coplanarity error for each lead; and

comparing each of the differences and the coplanarity errors to predetermined specification limits to determine acceptability of the semiconductor package.

13. The method of claim 9 further comprising identifying an anomalous shape which cannot be correlated to the stored predetermined direction edge shape and dilating the anomalous shape.